Oilfield Review Summer 2008

Summer 2008

Multizone Fracturing

Groundwater Management

Triaxial Induction Resistivity

Oilfield Review

SCHLUMBERGER OILFIELD REVIEW

SUMM

ER 2008VOLUM

E 20 NUM

BER 2

21502schD1R1.qxp:21502schD1R1 9/3/08 2:13 AM

08-OR-003-0

21502schD1R1.qxp:21502schD1R1 9/3/08 2:00 AM

« L’huile, l’huile » French engineer Gilbert Deschâtre wrotein November 1929: “Oil, oil.” Translated, his letter senthome to Paris from Seminole, Oklahoma, USA, continues,“If only we had a method to detect it, they would never findour prices too expensive. They are always asking, ‘Do youknow how to distinguish an oil sandstone from a watersandstone?’...Obviously we can tell the difference, providedthat they tell us first that the layer is a sandstone!” Thusdid Deschâtre report to the headquarters of Société deprospection électrique on trials of a new technology formeasuring electrical properties of rocks using instrumentshanging in a borehole.1

Two years earlier, a young engineer named Henri-GeorgesDoll had performed the first “electrical coring” experimentin Pechelbronn, France, implementing an idea conceivedby Conrad Schlumberger and his brother Marcel. Dolljoined Deschâtre on the cold plains of the Greater Seminoleoil field for the field tests of what came to be called elec-trical well logging. This new measurement transformed thesearch for oil with a simple, but profound observation:water is a conductor; oil is an insulator. Refined over theyears, this basic observation is still vital to the industry.

The first electrical logging method led to several genera-tions of electrode tools that measure local resistance tocurrent flow (“resistivity”) by injecting steady electricalcurrent into the earth. The latest tool in this line, the RtScanner triaxial induction tool, is descended directly froma breakthrough device that Doll invented in the late 1940s.Using principles of electromagnetic induction, his designcreates alternating current in rocks and thereby solves theproblem of remotely sensing the resistivity outside wellsdrilled with oil-base mud, which does not allow the pas-sage of direct current (see “Triaxial Induction—A NewAngle for an Old Measurement,” page 64).

Deschâtre’s call still rings out today, but with a 21st cen-tury echo: “Bypassed oil!” Or, even more simply: “Reserves!Reserves!” As demand for oil puts pressure on its supply—and its price—today’s industry requires increasingly bet-ter estimates of the total amount of oil in place, or still tobe recovered, in smaller, deeper and more complex reser-voirs. To do this requires formation evaluation that solvesat least three separate puzzles: lithology, geometry andsaturation. The most important of these is saturation—thepercentage of hydrocarbon in the pores—but its quantifi-cation depends on answering the first two.

Inducing the Next Developments

Lithology: What types of rocks are present? Geologiststoday can confidently piece together the type andcomposition of a reservoir rock, along with its porosity,from modern nuclear, acoustic and magnetic resonancelogs—or from core samples, if all else fails.

Geometry: What is the rock’s configuration? The puzzleof complex geometry, especially the 3D reservoir geometryaround horizontal wells, remains stubbornly fragmented.Now, the Rt Scanner tool attacks the problem of geometryhead on: it is the first truly 3D resistivity tool, capable ofinducing and measuring the effects of electrical currentscirculating in any direction within the earth. Its first majorapplication will be in laminated reservoirs comprising thinoil sands interlaced with even finer shale layers. By mea-suring the resistances felt by currents flowing both paral-lel and perpendicular to the layering—which can differ bya factor of 10 or more—the new tool can provide a moreaccurate tally of the total oil present.

Development of the new induction tool was made possi-ble by impressive advances in electronics and in materials,allowing the creation of flexible, compact triaxial induc-tion coils. But above all, its conception and developmentrelied on new capability to model the device and its elec-tromagnetic field in three dimensions.

The challenge now lies squarely in the realm of inver-sion: in the hands of the theorists and modelers who mustconfront the problem of turning raw physical measure-ments into quantitative rock properties. The holy grail is amodel that makes no assumptions about how rocks and flu-ids are distributed around the well, but rather determinesthe geometry from the measurements. This is still a distantgoal, but finally, the Rt Scanner tool provides the informa-tion required to bring it into focus.

Michael OristaglioTechnology AdvisorSchlumberger-Doll ResearchCambridge, Massachusetts, USA

Michael Oristaglio, Technology Advisor to the Schlumberger Mergers andAcquisitions team, works at Schlumberger-Doll Research (SDR) in Cambridge,Massachusetts, and is responsible for scouting firms developing early-phasetechnology for the energy industry. He joined Schlumberger in 1982 in theMechanics Electrical department of SDR and has worked as a scientist andmanager in areas of seismic exploration, software development, electromag-netics, and technical communities. From 2000 to 2004, Michael worked at Witten Technologies, a small company developing ground-penetrating radarfor mapping underground utility networks. He received a combined BS and MSdegree in geology and geophysics from Yale University, New Haven, Connecti-cut, USA; an MS degree in geochemistry from University of Oxford, England;and a PhD degree in geophysics, also from Oxford.

1

1. On one point, Deschâtre was wrong: When he and Henri-Georges Doll metwith the geologist who had organized the logging trials, they learned that theclient was delighted with the results and eager to continue the work, but only ifthe cost of the service and crew was reduced from $2,000 a month to about $400.

21502schD2R1.qxp:21502schD2R1 8/30/08 3:00 AM Page 1

Schlumberger

Oilfield Review4 The Right Treatment for the Right Reservoir

In recent years, operators have looked to increased reservoircontact as a way to drain their reservoirs more efficiently.Hydraulic fracturing is often a key component of this strategybut has proved uneconomic when used with certain types ofcompletions. This article describes efforts to overcome thatfinancial challenge through efficiency-enhancing fracturingtechniques and tools.

18 Managing a Precious Resource

Groundwater constitutes a predominant share of the Earth’ssupply of drinkable water. Characterization of subsurfaceaquifers is essential for managing and sustaining our supply offresh water. Advanced logging, sampling and modeling tech-niques—some borrowed or modified from established oilfieldapplications—are proving vital for evaluating and managingthis precious resource.

34 Sand Injectites

Under certain conditions, unconsolidated sand can be remobi-lized and forced upward through overlying impermeable layers.These injected sands, called injectites, can form excellent payzones and enhance reservoir connectivity. Geologists and otherE&P professionals are using surface exposures of injectites,along with core, borehole images and surface seismic data, tounderstand the shape and distribution of injectites in the sub-surface. Examples from the USA, the North Sea and the Gulf ofGuinea show what injectites look like and how they can affectreservoir development.

Executive EditorMark A. Andersen

Advisory EditorLisa Stewart

EditorsMatt VarhaugRick von FlaternVladislav GlyanchenkoTony Smithson

Contributing EditorsRana RottenbergGlenda de LunaJudy JonesDavid Allan

Design/ProductionHerring DesignSteve Freeman

IllustrationTom McNeffMike MessingerGeorge Stewart

PrintingWetmore Printing CompanyCurtis Weeks

Address editorial correspondence to:Oilfield Review5599 San Felipe Houston, Texas 77056 USA(1) 713-513-1194Fax: (1) 713-513-2057E-mail: [email protected]

Address distribution inquiries to:Tony SmithsonOilfield Review12149 Lakeview Manor Dr.Northport, Alabama 35475 USA(1) 832-886-5217Fax: (1) 281-285-0065E-mail: [email protected]

Useful links:

Schlumbergerwww.slb.com

Oilfield Review Archivewww.slb.com/oilfieldreview

Oilfield Glossarywww.glossary.oilfield.slb.com

On the cover:

In this outcrop near Santa Cruz, California,USA, a dark sandstone formed by sedi-ment remobilization is encased in lightermudstone. The sand, sourced from a par-ent bed deeper in the section, traveledto the surface through dikes, then filledwith hydrocarbon, giving it a dark color.The dark layer continues in the back-ground. The inset at left depicts thegeometry of a hypothetical sand injec-tion. On the right, an inset shows a newtriaxial induction tool, used to calculatehorizontal and vertical resistivities andstructural dip at any well deviation.

2


Summer 2008Volume 20Number 2

85 Contributors

88 New Books and Coming in Oilfield Review

3

50 Moving Natural Gas Across Oceans

Companies with reserves of natural gas in remote locations cannow move it to consumers in distant markets. This gas, onceconsidered stranded, is liquefied and transported as liquefiednatural gas (LNG) in large, custom-designed vessels. New tech-nology is transforming every link in the LNG chain, from highlyefficient liquefaction plants to new, offshore import terminals.

64 Triaxial Induction—A New Angle for anOld Measurement

A new triaxial induction resistivity service overcomes many ofthe shortcomings of previous-generation induction tools.Although resistivity logs have been used to identify oil and gasdeposits for more than 80 years, it is now possible to properlyevaluate electrically anisotropic reservoirs and more accuratelymeasure resistivity in high-angle wells. In addition, high-qualitystructural dipmeter data far from the wellbore are a direct out-put with this new service.

Abdulla I. Al-KubaisySaudi AramcoRas Tanura, Saudi Arabia

Dilip M. KaleONGC Energy CentreNew Delhi, India

Roland HampWoodside Energy, Ltd.Perth, Australia

George KingRimrock Energy LLCDenver, Colorado, USA

Eteng A. SalamPERTAMINAJakarta, Indonesia

Jacques Braile SaliésPetrobrasHouston, Texas, USA

Richard WoodhouseIndependent consultantSurrey, England

Advisory Panel

Oilfield Review subscriptions are available from:Oilfield Review ServicesBarbour Square, High StreetTattenhall, Chester CH3 9RF England(44) 1829-770569Fax: (44) 1829-771354E-mail: [email protected] subscriptions, including postage,are 200.00 US dollars, subject toexchange-rate fluctuations.

Oilfield Review is published quarterly bySchlumberger to communicate technicaladvances in finding and producing hydro-carbons to oilfield professionals. OilfieldReview is distributed by Schlumberger toits employees and clients. Oilfield Reviewis printed in the USA.

Contributors listed with only geographiclocation are employees of Schlumbergeror its affiliates.

© 2008 Schlumberger. All rights reserved.No part of this publication may be repro-duced, stored in a retrieval system ortransmitted in any form or by any means,electronic, mechanical, photocopying,recording or otherwise without the priorwritten permission of the publisher.


4 Oilfield Review

The Right Treatment for the Right Reservoir

Bader Al-MatarMajdi Al-MutawaMuhammad AslamMohammad DashtiJitendra SharmaKuwait Oil CompanyAhmadi, Kuwait

Byung O. LeeJ. Ricardo SolaresSaudi AramcoUdhailiyah, Saudi Arabia

Tom S. NemecGoodrich PetroleumHouston, Texas, USA

Jason SwarenSugar Land, Texas

Loris Tealdi Eni Congo S.A.Pointe Noire, Republic of the Congo

For help in preparation of this article, thanks to MichaelDardis, Longview, Texas; Phil Duda and Donald Smith, Houston;Matt Gillard, Moscow; Shrihari Kelkar, Al-Khobar, Saudi Arabia;Hai Liu, Ahmadi, Kuwait; and Brad Malone, Pointe-Noire, Congo.AbrasiFRAC, ABRASIJET, ACTive, CoilFRAC, Contact,DeepSTIM, DivertaMAX, InterACT, PCM, POD, RapidSTIM,StageFRAC, StimMAP, SuperX, SXE, VDA (ViscoelasticDiverting Acid) and VSI (Versatile Seismic Imager) are marksof Schlumberger.PerfFRAC is a mark of Schlumberger, technology licensedfrom ExxonMobil Upstream Research Company.

Flow rates in most wells increase significantly after hydraulic fracture treatments.

In some completion configurations—notably multizone and long-reach, high-angle

wells—capital and operating expenses often negate economic gains from improved

ultimate recovery or accelerated production. This drawback is being addressed today

by combining more efficient, multizone fracturing tools and services with real-time

monitoring capabilities.

Among the strategies used today to produce alarger proportion of original reserves in place arehigh-angle, extended-reach wells, multizone wellsand recompletions aimed at capturing previouslyuneconomic or stranded oil and gas deposits.Recent improvements in geosteering technologyallow high-angle wells to be drilled to increasinglygreater distances, while preventing the well pathfrom straying beyond the upper and lowerboundaries of the pay zone. The result is an ability

to greatly increase wellbore contact with theformation and so significantly improve drainage.

Increased formation contact is essential tothe success of many of these long lateral wells.This is because most of the formations thatreadily give up hydrocarbons were discoveredand developed years ago. That leaves today’soperators the task of producing oil and gas fromunconventional or low-permeability sources suchas shale or the outer reaches of mature fields

> Improving formation contact in vertical and horizontal wells. A 100-ft [31-m], 81⁄2-in., vertical wellboreresults in about 222 ft2 [20.6 m2] of formation contact (far left ). A 2,000-ft [610-m], 81⁄2-in., horizontalwellbore drilled in the formation increases formation contact 20 times compared to the 100-ft verticalwell (middle left ). A 150-ft [45-m] fracture length in the vertical well increases formation contact 270times over that of the untreated vertical well and 13.5 times that of the 2,000-ft untreated horizontalwell (middle right ). When the 2,000-ft horizontal well is treated with ten, 75-ft [23-m] long fractures,formation contact increases to 1,013 times that of the untreated vertical well and 50 times that of theuntreated horizontal well (far right ).

100-ft untreatedvertical well

222 ft2

of contact

2,000-ft untreatedhorizontal well

20 x vertical

100-ft vertical well treatedwith 150-ft fracture

270 x vertical13.5 x horizontal

2,000-ft horizontal well treatedwith ten 75-ft fractures

1,013 x vertical50 x horizontal


Summer 2008 5

where reservoir quality may be low.1 Whileextended-reach wells play a significant role inimproving reservoir contact, exposure to theformation can be further increased withhydraulic fractures.

In vertical wells, hydraulic fracturing canimprove reservoir contact several hundredfold.In horizontal wells, the improvement isexponential (previous page).2 While the resultsfrom fracturing high-angle, long-reach wells

have been encouraging, many of thesetreatments often fail to deliver expectedeconomic or production gains. This outcome is afunction of the original completion methods: tomaximize wellbore-formation contact, thesewells are traditionally completed open hole orwith slotted or preper forated liners across theproduction zone.

In an openhole completion, effective stimu -lation along the horizontal wellbore is almost

impossible using traditional bullheadingmethods, since it is difficult to place fracturefluids and acids precisely within the formation.Typically, with the use of standard methods, onlythe upper sections, or heel, of the wellbore aretreated, with little fluid ever reaching the middleor lower intervals (left).3

When operators choose to completehorizontal wells with cemented liners, individualzones can be more readily isolated and treated.However, as with any multizone treatment, thecosts of multiple, time-consuming trips per zoneoften outweigh the value of the resultingincreased production.

Despite these hurdles, and because hydraulicfracturing does consistently result in increasedproduction, demand continues to grow for thepractice in all types of wells. In an effort to attainbetter results—from both a cost and productionstandpoint—service companies are delivering

1. For more on production from gas shale: Boyer C,Kieschnick J, Lewis RE, Suarez-Rivera R and Waters G:“Producing Gas from Its Source,” Oilfield Review 18,no. 3 (Autumn 2006): 36–49.

2. Chariag B: “Maximize Reservoir Contact,” Hart’s E&PMagazine (January 2007): 11–12.

3. Al-Naimi KM, Lee BO, Bartko KM, Kelkar SK, Shaheen M,Al-Jalal Z and Johnston B: “Application of a Novel Open-Hole Horizontal Well Completion in Saudi Arabia,”paper SPE 113553, presented at the SPE Indian Oil andGas Technical Conference and Exhibition, Mumbai,March 4–6, 2008.

> Fracturing openhole, horizontal wells. Microseismic events were captured during the treatment of ahorizontal, openhole completion in a Barnett Shale well. The multicomponent sensors in monitoringwells (green) indicate that almost all of the bullheaded treatment was absorbed in the heel, or uppersection, of the well (blue). As a result, the majority of the lower Barnett formation remained untreated.

8,500

8,000

7,500

7,000

6,500

–500 0 500 1,000 1,500

Distance along well, ft

2,000 2,500 3,000 3,500

Dept

h, ft

Lower Barnett

Treatment well

Perforations

21502schD4R1:21502schD4R1 8/25/08 6:47 AM Page 5

fracturing systems that can access, stimulate and isolate numerous zones in cemented and openhole, extended-reach and verticalcompletions, in a single intervention.

Some of these multizone fracturing tech -niques are also designed to solve issues ofdegraded tubulars and control of fracture fluidplacement through the use of conveyance systemssuch as coiled tubing (CT) or jointed pipe.4

This article describes fracturing and comple -tion systems that allow operators to overcomethe economic and technological barriers tohydraulic fracturing in certain, increasinglyprevalent well types. Examples from NorthAmerica, Africa, Saudi Arabia and Kuwaitdemonstrate how these innovative approacheshave resulted in efficient, economically viablefracturing and acidizing treatments in reservoirsonce deemed poor candidates for such proce -dures. We will also examine recent innovationsthat allow operators to model and then monitorand refine their fracture treatments in real time.

An Eye on GrowthAs in many oil and gas operations, integrationwith real-time monitoring has greatly enhancedthe effectiveness of hydraulic fracturing. In thepast, downhole pressures were derived frommeasurements taken at the surface and extrap -olated to bottomhole conditions. Today, thesemeasurements are acquired directly at thesandface in real time using coiled tubing with an installed fiber-optic line (left). Themeasurements are obtained successfully despitethe extremely harsh downhole environmentcreated during hydraulic fracturing operations.

Fiber-optic-equipped coiled tubing (FOCT)systems feature a downhole sensor package thatsends real-time downhole depth, temperatureand pressure data to the surface. In addition, theoptical fiber can obtain distributed temperaturereadings at 3-ft [1-m] intervals. Data aretransmitted from the toolstring through thefiber-optic line to an electronics package thatconverts the fiber-optic signal to a wirelesssignal. This, in turn, allows transmission of thedata to a control cab where the information canbe viewed remotely through a command-and-acquisition software program.5

Operators also can gain considerable valuefrom accurate definition of the fracture systemgeometry as it is being created. Equipped withsuch knowledge, engineers can design successivefracturing operations within a field to avoidundesired results. In the past, fracture mappingwas performed through postfracture analysismeasurements such as temperature logs, radio -active tracers and tiltmeter surveys. However,these tools have shortcomings. Temperature logsor radioactive tracers can provide only near-wellbore fracture height and width. And whileinformation about the azimuth and symmetry ofthe fracture may be gained from surface anddownhole tiltmeter mapping, these methods

do not accurately assess the fracture’s height,length and width.6

More recently, service companies havedeveloped the ability to describe fracturegeometry using borehole seismic methods.7 TheStimMAP hydraulic fracture stimulation diag -nostics service uses multicomponent receivers inan offset well to record the microseismic activitycaused by the creation of hydraulic fractures inthe treated well. To create the velocity modelneeded for microseismic data analysis andprocessing, a seismically calibrated velocitymodel survey is performed in a nearby moni -toring wellbore. This borehole seismic survey isperformed before fracturing.

The map of these microseismic events allowsengineers to better understand the developmentof induced fractures in time and space. Mappingalso provides valuable geological insight into thetreated formation.8

Engineers at the monitoring or treatmentwell can communicate with other locations usingthe InterACT connectivity, collaboration, andinformation system. Remote office locations canbe included in the communications loop forinstant processing and interpretation of the data.

The StimMAP system uses real-time data tolocate microseismic events automatically in 3D space (next page).9 Comparing the fracturemapped by the StimMAP service with a fracturingdesign and evaluation software model providesuseful information for improving future treat -ments. The lessons learned enable operators tooptimize well-stimulation costs and gain insightfor new in-field drilling opportunities.

The StimMAP system was recently applied ina multizone fracturing operation in an east Texashorizontal well. During stimulation aimed at thethird zone, engineers observed unintendedmicro seismic activity in the region of what was to be a fifth zone. Following unsuccessfulattempts to redirect the fracture, the companyhalted operations.

Engineers coupled StimMAP Live services—aspecific application of the StimMAP system thatpermits engineers to monitor and, if necessary,alter fracture treatments as they are beingperformed—with pumping data to diagnose themechanical problems that were causing thefracture to veer from its planned direction. Thejob was then resumed and three more zonestreated. Without the insight afforded engineersby real-time feedback, six fracture treatmentswould have been pumped into the same zone.Instead, the operator was able to salvage three of

6 Oilfield Review

> Real-time data. Fiber-optic cable inside coiledtubing delivers a distributed temperature profilethroughout the wellbore. Temperature variationsprovide information that shows where fracturefluids are entering the reservoir. In relatively low-rate applications, ACTive in-well live performancemeasurements can be made as the fracturetreatment is pumped directly down the coiledtubing. During what are termed high-rate jobs—more than 450,000 lbm [204,117 kg] of sandpumped at 10 bbl/min [1.6 m3/min]—the fiberbundle, known as the tether, begins to moveforward and buckle, causing it to fail. For thosejobs, the system monitors conditions as the sandis pumped down the CT-casing annulus. Thesensor package shown here includes a battery(white), a circuit board with temperature sensor(green) and pressure transducers (light blue).


Summer 2008 7

the remaining five treatments while avoiding theweeks of diagnostic workover costs thatotherwise would have been necessary.

Making Multizone PayIn times of high commodity prices, operators arenaturally anxious to make the most of theirassets by producing as much hydrocarbon as is economically feasible. To do so, they often complete numerous zones with a singlewellbore or expose long intervals of formationthrough horizontal or high-angle drilling. Asdiscussed above, results from traditionalapproaches to fracturing these wells may fallshort of operator expectations for economic ortechnological reasons.

Multizone fracturing, as opposed to tradi -tional methods that include multiple trips perzone, targets both economic and techno logicalconcerns. Through efficient practices and newtechniques, these services may reduce weeks ofrig costs to a few days or even entirely eliminatethe need for a workover or drilling rig. Multizonefracturing practices also are able to deliver moreeffective treatments that optimize formationcontact because they can more accurately placetreatments without adding risk.

Service companies have customized systemsto address the varied types of multizone wellsoperators seek to treat. Schlumberger hascreated a four-category package of hydraulicfracturing services based on well type andoperator philosophy. Called Contact staged frac -turing and completion services, the categoriesinclude the following: • conventional systems that require separate

trips into the well to perforate a zone in onetrip, then stimulate and isolate it in a secondtrip, repeating that process for each zone

• intervention systems that perforate, fracture-stimulate and isolate numerous zones in asingle trip

• permanent systems that fracture-stimulateand isolate multiple zones in one pumpingoperation using assemblies that remain in thewell as part of the completion

• dynamic systems that use a degradable divert-ing material to successively plug and isolatetreated perforations and divert stimulations toother intervals in a continuous operation.

Conventional fracturing—pumping thefracture fluid and proppant or acid down thecasing or fracture workstring—is most effectivefor massive hydraulic fracture treatments inwhich hundreds of thousands of pounds of sandare pumped downhole at high rates. In cased

holes, the reservoir is accessed throughperforations created by wireline, abrasive jettingor shifting sleeves in the workstring.

4. For more on coiled tubing stimulation: Degenhardt KF,Stevenson J, Gale B, Gonzalez D, Hall S, Marsh J andZemlak W: “Isolate and Stimulate Individual Pay Zones,”Oilfield Review 13, no. 3 (Autumn 2001): 60–77.

5. Julian JY, West TL, Yeager KE, Mielke RL, Allely JN, Jenkins CN, Perius PD, Bucher RL, Foinquinos CI, Forcade KC, Fagnant JA, Montgomery DB, McInnis JGand Sack JK: “State-of-the-Art Coiled Tubing Operationsat Prudhoe Bay, Alaska,” paper IPTC 11533, presented atthe International Petroleum Technology Conference,Dubai, UAE, December 4–6, 2007.

6. Le Calvez JH, Klem RC, Bennett L, Erwemi A, Craven Mand Palacio JC: “Real-Time Microseismic Monitoring of Hydraulic Fracture Treatment: A Tool To Improve Completion and Reservoir Management,” paper SPE 106159, presented at the SPE Hydraulic FracturingTechnology Conference, College Station, Texas, January 29–31, 2007.

7. For more on borehole seismic applications: Blackburn J,Daniels J, Dingwall S, Hampden-Smith G, Leaney S,Le Calvez J, Nutt L, Minkiti H, Sanchez A and Schinelli M:“Borehole Seismic Surveys: Beyond the Vertical Profile,”Oilfield Review 19, no. 3 (Autumn 2007): 20–35.

8. Le Calvez et al, reference 6.9. For more on microseismic mapping: Blackburn et al,

reference 7.Bennett L, Le Calvez J, Sarver DR, Tanner K, Birk WS,Waters G, Drew J, Michaud G, Primiero P, Eisner L, Jones R, Leslie D, Williams MJ, Govenlock J, Klem RCand Tezuka K: “The Source for Hydraulic Fracture Characterization,” Oilfield Review 17, no. 4 (Winter 2005/2006): 42–57.

> Mapping the fracture. Shown here are a map view (left ) and a transverse view (right ) of microseismic locations from a four-stage slickwater stimulationtreatment in the Barnett Shale. StimMAP services were chosen to create an optimal fracture design using accurate image geometry of the hydraulicfracture as it developed. Microseismic data were acquired with the multishuttle VSI Versatile Seismic Imager and processed on location to generate a 3Dcomputer image of the fracture system. This allowed the stimulation treatment of subsequent stages to be reengineered.

5,000

5,000

5,250

5,500

5,750

6,000

6,250

6,500

6,750

7,000

7,250

7,500

7,750

4,500

4,000

3,500

3,000

2,500

Nor

th-s

outh

dis

tanc

e, f

t

True

ver

tica

l dep

th, f

tEast-west distance, ft Distance along projection, ft

–4,000 –3,500 –3,000 –2,500 –2,000 –1,500 –1,000 –500 0 500 –5,400 –4,900 –4,400 –3,900 –3,400 –2,900 –2,400 –1,900 –1,400 –900 –4001,000

2,000

1,500

1,000

500

0

–500

Bethany 1

Godley 1 Stage 4

Stage 3

Stage 2

Stage 1

Bethany 3H

Perforation

Godley 1

Bethany 1

Stage 4

Stage 3

Stage 2

Stage 1

Bethany 3H

Perforation

Estimated Lower Barnett top

Estimated Ellenberger top


When multiple intervals are open within asingle zone, diverting fluid from one to another inorder to treat each of them may be accomplishedthrough such practices as limited entry, ballsealers, chemical diversion, composite bridgeplugs and sand plugs. Limited entry is created byplacing fewer perforations across certainsections to increase friction at the openperforations. This diverts the fluids from a zone

that, because of high permeability or otherfactors, may have absorbed the bulk of thetreatment at the expense of other intervals orzones (left).10

Composite bridge plugs are isolation barriersset in the casing above the treated zone anddrilled out later, usually in a separate workstringdrilling operation. This incurs a time penalty andadds operational risk. Additionally, the timebetween treating the lowest formation andflowing it back can sometimes be measured inweeks; in some cases, that may be enough timefor the fluids to leave residue in pore spaces,causing significant damage to the formation.

In wells with openhole completions andunconsolidated formations, conventional frac -turing may include deploying a completionstring—typically a slotted or perforated liner—to ensure wellbore integrity. The entire well maybe fracture-stimulated by pumping treatmentfluid down the casing or fracturing string andinto the formation in a practice known asbullheading. As in cemented completions, once acompletion string is in place, diversion may beattempted by using limited entry, ball sealers ortraditional chemical diversion.

Staging an InterventionThe intervention category of hydraulic fracturingcomprises three services: AbrasiFRAC abrasiveperforating and fracturing service, PerfFRACselective perforating, fracturing, and stageisolation with ball sealers, and CoilFRACstimulation through coiled tubing.

The AbrasiFRAC technique enables accurateplacement of fracturing treatments down thecasing or the workstring-casing annulus. It also

reduces near-wellbore pressure drop from thewellbore to the reservoir, which decreases thefrequency of near-wellbore screenouts whenproppant stops entering the formation and buildsup inside the casing. The technique isparticularly well-suited to treating formationswith high fracture-initiation pressure and areasin which precise placement is critical to thesuccess of the stimulation.

The system is based on a well-established oil industry technique for cutting casing andtubulars downhole: a slurry containing abrasivesolids is pumped at high differential pressuresthrough an ABRASIJET hydraulic pipe-cuttingand perforating service gun conveyed on a workstring. The resulting high-velocity fluid stream cuts through tubulars andsurrounding cement, then penetrates deep intothe formation (below).

The cutting tool is used to perforate thecasing and formation. The abrasive material isusually 20/40- or 100-mesh fracturing sand,which is compatible with the speciallyengineered jet guns. Sand plugs may be used toprovide zonal isolation between the fracturingtreatment zones. The jet guns, which areavailable in various size and phase config -urations, also can be used with retrievable ormillable bridge plugs for isolation.

One example of the use of the AbrasiFRACservice was in the highly laminated Hosston sandsof the Sligo field in northern Louisiana, USA. TheHosston sands contain many thin, gas-bearingsands alternating with water-bearing sands, withvarying levels of depletion. Typically, wells in thisarea are completed with multistage stimulation

8 Oilfield Review

> Conventional fracture fluid diversion throughball sealers. Once the calculated amount oftreatment fluids is pumped into a targeted zone(tan), the flow is diverted to another (blackarrows). The most common method of diversioninvolves ball sealers (black), made of nylon, hard rubber, biodegradable collagen, or acombination of these, introduced into the slurryso that they reach the perforations at the end ofthe treatment. The balls create a seal across theperforations, causing the treatment to divert tothe next open set of perforations. By repeatingthis procedure, numerous intervals per stagecan be treated without shutting down pumps orsetting plugs.

> High-pressure perforating and treatment. The AbrasiFRAC service uses a high-performance abrasive jetting tool to operate continuously under harshdownhole conditions. After depth correlation, abrasive sand slurry is pumped through the nozzles. The resulting high-velocity fluid stream perforatestubulars and surrounding cement, then penetrates deep into the formation (far left). The zone is then treated (middle left) and the cutting tool is pulled upthe hole to the next zone. Once the first treatment is complete, a sand plug may be set for isolation and the next zone perforated (middle right). Thissequence can be repeated as often as necessary in a single operation. When all zones are treated (far right), the sand plugs are reverse-circulated out.

Cutting perforations Pumping fracture treatmentdown the annulus

Cutting next set of perforationsafter placing sand plug for isolation

Reversing out sand plugafter fracturing treatment


Summer 2008 9

treatments using energized fracturing fluids andbridge plugs for isolation between zones.

In an effort to improve cost and timeefficiencies, EOG Resources field-tested theAbrasiFRAC service. The technology allowed theoperator to stimulate multiple intervals within awell in a single field operation and to moreeffectively and efficiently stimulate theindividual sands. Treatments varied from four tonine stages using CO2-energized fracturingfluids. The result was to cut water production by85% while doubling gas production.

Another approach to efficiency is to treatzones immediately after perforation without firstpulling the guns out of the hole. This step alonesaves one running and one pulling trip per zone.The PerfFRAC service is designed to performhigh-rate treatments down the casing while theperforating gun assembly remains in the well -bore. First, the guns for each zone are run in thehole and the first zone is perforated. Then, as thefirst zone is being treated, the unfired guns aremoved up the borehole and positioned to shootthe holes for the second zone.

At the end of the first zone’s treatment, ballsealers in a fiber diverting fluid are pumped intothe well. A rise in pump pressure indicates thatthe ball sealers and slurry have sealed againstthe perforations of the treated zones. At thatpoint, the guns for the second zone are fired, andthe second treatment, again tailed in with ballsealers and fiber-infused diverting fluid, ispumped. This process is repeated for multiplezones. To date, as many as eight sets of guns havebeen run and six separate zones treated in asingle intervention (below).

The PerfFRAC service often results in betterproduction rates than other, less efficienttreatments because it allows precise targeting oftreatments, ensuring that no zones are under -served. The method also allows the well to beflowed back immediately and so avoids the risksassociated with milling out composite bridgeplugs and leaving potentially damaging fluids inthe formation for an extended time period.

Goodrich Petroleum was seeking to bring justsuch efficiencies to its Cotton Valley field in eastTexas. Engineers had been treating these wells

using traditional practices: perforate the firstzone, fracture-stimulate it, flow the well to cleanit up, and finally set a composite bridge plug for isolation.

This process was repeated for each zone ofinterest. Once the last zone was treated, a coiledtubing unit was brought on location to drill outthe composite bridge plugs. This sequence costthe company US $250,000 and took five days tocomplete. Goodrich opted to use the PerfFRACservice to complete 23 of its wells and in doing soreduced the five-day operation to one day, whileeliminating the need for bridge plugs and coiledtubing milling.

But the break with past practices paid a moreimportant dividend than just lower operatingcosts and shorter time to production. In the first

> Multizone treatments in vertical wells. Multiple stacked zones in vertical wells may be stimulated using limited entry. The PerfFRAC technique shownhere uses high-rate fracture stimulation treatments down the casing with a perforating assembly in the wellbore. Once the lower zone is perforated (left ),the guns are moved up the hole and positioned for the next set of perforations; then the perforated zone is fractured (middle). Ball sealers (green) isolatethe treated zone and the next zone is perforated (right ).

10. For more on diversion techniques: Samuel M and Sengul M: “Stimulate the Flow,” Middle East & AsiaReservoir Review no. 4 (2003): 42–55.


10 Oilfield Review

> Multistage versus traditional fracturing practices. In the Cotton Valley field of Texas, Goodrich Petroleum used multistage fracturing to reduce treatmenttime (blue) from five days to one and costs (green) from US $255,000 to US $155,000 per well (left ). In a 23-well program in the same field, the operatorproduced 25 MMcf more gas compared with what would be expected following traditional treatment methods that often sacrifice treatment efficiency tomeet economic goals (right, horizontal scale not linear).

Conventional methods

Cos

t, U

S d

olla

rs

Tim

e to

com

plet

ion,

day

s

0

40,000

80,000

120,000

160,000

200,000

240,000

280,000

0

1

2

3

4

5

6

7

PerfFRAC service

$255,000

$155,000

5 days

1 day

Time

Cost

145

Cum

ulat

ive

aver

age

gas

prod

ucti

on p

er w

ell,

Mcf

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

Time, days of production

9030 18060

25 MMcf


PerfFRAC service

> Workstring-conveyed selectivity. By combining a workstring with selective fracturing technology, operators can treat multiple zones in a single trip. In newwells, each zone is perforated conventionally in one wellsite visit. Coiled tubing or jointed pipe is then deployed into the wellbore with an openhole packerbottomhole assembly (right ). The bottom zone is isolated by packers above and below the target formation, and the fracture stimulation is pumped throughthe workstring (left ). Residual proppant is reverse-circulated out of the wellbore and the packer is moved to the next zone, where the process is repeated.The insets represent real-time monitoring of each treatment.

CT connector

CT disconnect

Top cup

Bottom cup

Bottom plug

Spacer

Fracturing sub

Coiled tubing


Summer 2008 11

180 days of production, the 23 wells recovered anadditional 25 MMcf [708,000 m3] of gas—a 22%increase over wells completed using conven -tional methods. This gain allowed the operator toincrease its estimated ultimate recovery per wellby 10% (previous page, top).

In all, using the PerfFRAC system savedGoodrich Petroleum 92 completion days on23 wells. Additionally, reduction in equipment onlocation saved another 25% on total completioncosts per well. Gas-to-market time was reducedby four days per well, netting the operator anadditional 6 MMcf [169,900 m3] of initial gas.

Diversionary TacticsWith the adoption of ball sealers and limitedentry, treated zones can be isolated and thefracture treatment diverted to untreated zones.While these isolation and diversion techniqueshave the advantage of significantly reducing thenumber of trips and the costs required tofracture wells with multiple zones, because ofdifferences in fracture gradients between sandswithin a well, these methods leave some zonesineffectively treated.

One solution to this shortcoming is to isolateand stimulate each zone individually with atreatment designed specifically for its param -eters. The trick is to do so without sacrificing theefficiencies gained from other practices such aslimited entry and ball sealing. To that end,

engineers have developed systems that isolatezones between sealing elements by usingopenhole packers that may be set, unset andreset numerous times.

The CoilFRAC coiled tubing stimulationservice uses an openhole packer assemblydeployed on a workstring across the bottom zoneafter the entire well has been perforatedconventionally (previous page, bottom). Thestimulation fluid is then pumped down thetubing string, through the treatment sub of thepacker tool and into the isolated interval.Residual proppant is then reverse-circulated outand the packer moved to the next zone. Thismethod not only permits stimulation of all zonesin a single intervention but, like other Contactservices, increases treatment efficacy byallowing the operator to customize eachtreatment to suit each zone.

In older wells, this type of service is especiallysuited for accessing bypassed reserves and forrefracturing previously completed zones.11 In thisapplication, the goal is not only to minimize thecost of fracturing mature assets, but to do so whileprotecting downgraded casing from high treat -ment pressures and abrasive, proppant-ladenfluids. Using a workstring as a conduit offers theadded advantage of allowing the operator to treattargeted zones without first having to kill thewell—a procedure from which older, pressure-depleted formations may not recover.

The value of customizing discrete treatmentsto suit the needs of each interval in a multizonewell was clearly demonstrated in a RockyMountain oil field of the USA. The field containsmultiple vertical sand layers that are from 5 to60 ft [1.5 to 18.3 m] thick and distributed from adepth of 2,000 ft to 5,000 ft [609 to 1,524 m]. Theoperator had been completing wells in this fieldprimarily with limited entry, but had used bridgeplugs when the distance between layers wassignificant. However, because of the variedfracture gradients exhibited in each sand layer,many zones were not being effectively stimu -lated. In addition, some marginal sand layerswere left untreated for economic reasons.

In its search for an effective way to stimulateeach zone without increasing completion costs,the field’s operator chose to use the CoilFRACsystem. The decision paid off. For example, onewell in the field had been producing 1.9 MMcf/d[53,800 m3/d] from one limited-entry fracturestimulation of multiple layers. With the packerassembly, bypassed layers were perforated andthe entire well was restimulated. Eight fracturestimulations were performed in one day, and thestabilized production rate from the well wasrecorded at 5.3 MMcf/d [150,100 m3/d].

In addition to a more effective stimulation ofeach layer, the treatments took only one to twodays, compared with the several weeks requiredto do the jobs using conventional techniques. Inthe four-well test, average production rates of thewells with the CoilFRAC system were more thantwice those of standard completions (above left).As a consequence, recoverable reserves per wellwere increased by more than 75%.

In another example, this time in southeasternAlberta, Canada, engineers were experiencingsimilar setbacks in their stimulation efforts in ashallow gas field. The wells in the area areusually completed in four zones, ranging from820 to 1,480 ft [250 to 451 m] deep. The forma -tions are composed of layered, silty sands thatfracture easily.

Historically, operators in this area have usedvarious fracturing techniques, including com -posite bridge-plug isolation. On a four-zonecompletion, such conventional practices requireeight trips into the wellbore and an additional16 days to complete the well with flowbackrequired between each treatment.

11. For more on refracturing operations: Dozier G, Elbel J,Fielder E, Hoover R, Lemp S, Reeves S, Siebrits E, Wisler D and Wolhart H: “Refracturing Works,” Oilfield Review 15, no. 3 (Autumn 2003): 38–53.

> Increased production rates. In one Rocky Mountain gas field, the average production rate fromwells treated with the CoilFRAC system was more than twice that of wells in the same field treatedconventionally.

0 1 2 3 4 5 6 7

2,500

2,000

1,500

1,000

500

Time, months

+1,000 Mcf/d

Aver

age

prod

uctio

n M

cf/d

One-year average production of 4 wells receiving CoilFRAC treatmentFour-year average production of 14 wells receiving conventional treatment


The more efficient, commingled multizonestimulations that rely on limited entry or ballsealers for diversion gain time compared withtraditional practices but, as discussed earlier,sacrifice production to do so. In addition, wells inthe area are generally stimulated in groups to makethe most efficient use of fracturing equipment.

The operator determined that, to optimizestimulation and production, each zone should befractured individually at reduced pump rates.That decision made CT-conveyed fracturingtreatments an obvious choice and resulted inreducing operation time to the point that crewswere treating two wells per day (left). This was acritical advantage because for environ mental andeconomical reasons, summer access is restrictedto 10- to 14-day operational win dows. In all, usingthe CoilFRAC service saved the company US$110,000 while increasing production by morethan 190% in one field.

Permanent SolutionsMotivated by a desire to minimize the number ofinterventions or tools introduced into horizontaland high-angle wells, some operators prefer totreat zones using equipment that will becomepart of the permanent completion design. Oneway to do so is to complete the well usingconventional casing with sliding sleeves. Inopenhole completions, such a system includeshydraulically operated openhole packers tocreate a seal against the wellbore wall.

In either cemented or openhole wells, eachzone is treated through the sliding sleeves. Thepurposes of the cement and openhole packersare the same: to provide isolation between zonesof different treating pressures and to ensuretreatment along the entire length of the well.

Like other multizone services, permanentsolutions also reduce risk by limiting trips in thehole to set and remove bridge plugs andincrease treatment efficiency by allowing theoperator to design each treatment for a specificzone. This strategy also increases the number ofzones that can be treated because operators

12 Oilfield Review

12. Seale R, Donaldson J and Athans J: “Multistage Fracturing System: Improving Operational Efficiency and Production,” paper SPE 104557, presented at the SPE Eastern Regional Meeting, Canton, Ohio, USA, October 11–13, 2006.

13. Skin is a dimensionless factor calculated to determinethe production efficiency of a well by comparing actualconditions with theoretical or ideal conditions. A positiveskin value indicates some damage or influences that are impairing well productivity. These are the result ofcompletion or drilling fluid residue left on or in the formation. A negative skin value indicates enhanced productivity, typically resulting from stimulation.

> Cost, time and production. In shallow wells in southeastern Alberta, theefficiency gains of workstring-conveyed fracturing over traditionaltechniques are evident by all measures. Standard methods (blue) consumedUS $400,000 and 16 days per well, resulting in a 16% production gain. In thesame field, the CoilFRAC system (pink) cost US $290,000 and took 4 days,resulting in a 192% increase in production.


CoilFRAC service

$400,000

$290,000

16 days

Increased production

4 days

16%

192%

Cost Time

> Permanent service. Multiple fracture stages can be completed in a single trip by isolating the targetformation between openhole packers. The treatment fluid is delivered through frac ports in the tubularsection between the packers (top). During StageFRAC service operations (bottom), a ball (red),pumped down the wellbore with the tail end of the treatment fluid, lands in a seat inside a slidingsleeve. The resulting pressure buildup forces the sleeve to the open position. Fluid is then forced toenter the interval above the landed ball and the seat. At the same time, the ball and seat form a sealthat acts as a plug to isolate the lower, previously treated zone. By using progressively larger seat andball diameters from the deepest zone to the shallowest, the entire formation may be treated uniformlyin a single intervention.

Frac ports Openhole packers


Summer 2008 13

often address risk by reducing the number ofplug pulling and running trips. This inevitablyleads to limiting the number of zones that maybe isolated and treated.

If engineers choose at some point torefracture a conventionally completed well inwhich all perforations are left open, they must doso through the workstring. In addition to the costof the rig required to do that and the riskinvolved, fracturing through a workstringintroduces added frictional forces that limit theflow rate during fracturing, so the optimal designcannot be accomplished.12

StageFRAC and RapidSTIM multizonefracturing and completion services incorporatePackers Plus technology—openhole packers thatare run on conventional casing to segment thereservoir with ball-activated sleeves placedbetween each set of openhole packers. The twosystems are nearly identical except that theStageFRAC service treats the isolated zonesthrough frac ports, and the RapidSTIM servicedoes so through jets. Both frac ports and jets arelocated between the packers and behind ball-activated sliding sleeves (previous page, bottom).This mechanical diversion allows for precise fluidplacement, complete zonal coverage and greatereffective fracture conductivity.

Eni Congo turned to this solution when facedwith a significant challenge in its offshoreoperations near the Republic of the Congo coastin West Africa. These aging fields still containlarge quantities of reserves in low-permeability(10 mD), consolidated formations that are barelyeconomic to produce using conventional stimu -lation methods. Formerly, stimulation consistedof matrix acidizing to eliminate or reach slightlynegative skin.13

Eni chose the StageFRAC service for threeexisting cased and perforated wells in theKitina field where eight hydraulic, proppedfractures were placed in three recompleted,cased-hole wells. These jobs were being pumpedfrom a support barge to an offshore platformwith limited deck space. As a result, only twozones could be pumped before the vessel had tobe restocked.

The first well pumped was the KTM W6 ST(right). Two of the three zones treated down astimulation liner were pumped without shuttingdown the pumps. Once the bottom interval hadbeen treated, a ball was pumped, the zone wasisolated, and the next zone was opened. Thethird zone was treated separately.

> Offshore permanent solution. The KTM W6 ST wellbore has a 7-in.cemented liner run from 1,600 to 3,110 m [5,250 to 10,204 ft]. The well wasoriginally produced from a deeper interval and recompleted as shown. TheStageFRAC service was used to perforate the well in three intervals: 2,785 to2,810 m [9,138 to 9,220 ft], 2,820 to 2,865 m [9,252 to 9,400 ft] and 2,870 to2,910 m [9,416 to 9,548 ft]. New perforations were added in 2-m [6-ft] groupsto each of the existing perforation sets to reduce risk of screenout caused bytortuosity. The packer assembly, run on 41⁄2-in. tubing as a production liner,straddles the three perforated intervals.

7-in. x 4 1⁄2-in. cased-holehydraulic packer


4 1⁄2-in. hydraulic frac port

4 1⁄2-in. toe circulating sub

7-in. permanent packer

Perforations at2,785 m2,810 m

Frac port at2,801 m


Frac port at2,861 m


Frac port at2,909 m

Circulating subat 2,913 m

Tubing bottomat 2,916 m


3-in. ball

3 1⁄4-in. ball


Three zones were also treated in the secondof the three-well series. In the final well, it wasdetermined that the lowest zone was too close toa water contact and this zone was left untreated.Stabilized production from the three wellsincreased 230% over their previous performance.Production prior to fracturing was around590 bbl/d [94 m3/d]; after treatment, steady-stateproduction was 1,950 bbl/d [310 m3/d] (left).

This sequential approach holds specialpromise for the offshore arena wherecompleting a single fracture using conventionaltechniques can take a week and must be donefrom an offshore rig that costs several hundredthousand US dollars per day. By using openholepackers and sliding sleeves for isolation andfracturing the whole well in a single pumpoperation, each zone can receive extensivestimulation during a single mobilization of astimulation vessel (left).

On land, it is the ability of sequentialtreatments to effectively treat an entire wellcontaining numerous zones of differing qualitythat attracts operators to the technique. In SaudiArabia, operator Saudi Aramco had completed awell with a 5,000-ft openhole, hori zontal sectionthrough eight different zones of varyingpermeability. Because of its higher permeability,engineers expected most of the oil contributionto come from Zone 1.

Because of the length and trajectory of thehorizontal section, coiled tubing could not reachthe lower sections and so the plan was tobullhead an acid treatment. However, because ofits high permeability, Zone 1 at the heel of thewell took all the acid, and the other seven zoneswere left untreated. As predicted and because ithad received all the acid treatment, allproduction improvement came from Zone 1.

14 Oilfield Review

< Offshore capacity. The Schlumberger DeepSTIMgroup of stimulation vessels are specificallydesigned for fracturing and other fluid treatments.Their onsite quality control and mixing capacities(top), pump capacity (middle), and storagecapacity (bottom), and their dynamic positioningcapabilities render them self-sufficient. This elim-inates deck and storage space concerns onplatforms and the need for a costly offshore rig.Since these vessels are equipped to treat six ormore zones sequentially, they can do in six hourswhat requires six weeks by conventional means—a significant savings of time in view of offshorerig day rates.

Environmental wasteProppant to silo

12,000 lbm/min

(10,000 lbm/min

for DeepSTIM III)

Completion fluidto pumps To PCM gel hydration system

Three-compartmentmixer

Acid blender

0 to 30 bbl/min

0 to 70 bbl/min transfer rate

POD blender

Dual silo: 2,000 ft3

Acid lineLow-pressure line

High-pressure pumps

16,850 hydraulic horsepower (DeepSTIM)

21,450 hydraulic horsepower (DeepSTIM II)12,850 hydraulic horsepower (DeepSTIM III)

Dry bulk storage:

14,200 ft3 (DeepSTIM)

16,700 ft3 (DeepSTIM II)8,400 ft3 (DeepSTIM III)

Gel fluid capacity:

4,800 bbl (DeepSTIM)

6,600 bbl (DeepSTIM II)4,140 bbl (DeepSTIM III)

Completion fluid:

870 bbl (DeepSTIM)

860 bbl (DeepSTIM II)810 bbl (DeepSTIM III)

To wellhead0 to 50 bbl/min0 to 15,000 psi

High-pressure line

Bulk liquid additives Pressure-relief valve

Two 300-ft coflexip reels(1 for DeepSTIM III)

Raw acid capacity of 8,400 galUS(12,600 galUS for DeepSTIM III)

PCM gelhydration system

Acid tanks

Coflexip reels

Loading deckCrane

Silo

Liquid additives

Control room and laboratory

Satellite system

Wheel house

Iron racks

> Treatment results offshore Congo. Compared with production before stimulation, stabilizedproduction following the Kitina three-well fracturing campaign was increased by about 230%.

Well name Prefracturingoil rate, bbl/d

Initial postfracturingoil rate, bbl/d

Steady-statepostfracturingoil rate, bbl/d

Improved stabilizedproduction rate,

bbl/d

KTM-W6 (3 fracs)

160 2,100 600 + 440 bbl/d

KTM-111

(3 fracs)

300 900 650 + 350 bbl/d

KTM-107

(2 fracs)

130 1,000 700 + 570 bbl/d


Summer 2008 15

To remedy the problem, Saudi Aramcoengineers chose to use the StageFRAC service fora staged acid fracturing of all the zones. AlthoughZone 1 had been treated, it was decided to notopen it until the other seven were stimulated,cleaned up and tested. All zones were individ -ually stimulated in one pumping opera tion andimmediately flowed back. The seven previouslyuntreated zones were tested and the resultscompared with those in an offset well that hadbeen stimulated using coiled tubing and testedwith all zones open. The well that received theStageFRAC treatment had twice the productionof the offset well and triple its productivity index.

Saudi Aramco repeated this strategy in a field trial to hydraulically fracture longhorizontal, openhole wells in a deep, high-pressure, high-temperature Khuff carbonateformation. These wells, company officialsbelieved, were falling short of their productionpotential because of formation damage incurredduring drilling operations.

Engineers were also interested in thefeasibility of replacing unstimulated dual-lateralwells with hydraulically fractured single laterals.To that end it was imperative to employ a methodthat ensured stimulation of the entire length ofthe well—an impossible task in these long,complex completions using standard coiledtubing techniques.

The target of the field trial was the 3,835-ft[1,169-m] single lateral Well H-1 in the Khuffcarbonate. Three fracturing stages were planned.The first frac port was opened ahead of thetreatment by pressuring up on the workstring,and a step-rate injection test was performed toestablish fracture parameters. The first fracturestage was then bullheaded down the tubing at amaximum rate of 94 bbl/min [15 m3/min] and atreating pressure of 11,700 psi [80.66 MPa]. Atotal of 16,700 galUS [63.2 m3] of treatmentfluids, including acid and pad, were pumped.

A 2.5-in. ball was dropped, the second fracport was opened, and the second treatment of194,000 galUS [734.4 m3] of treatment fluid was pumped at 100 bbl/min [15.9 m3/min] and11,580 psi [79.84 MPa]. The final 243,222-bbl[38,648-m3] stage was pumped after a 2.75-in.ball was dropped and frac ports opened.

The well was cleaned up over a 54-hour flowperiod. Its performance was compared with thatof Wells H-2 and H-3, two offset nonstimulatedhorizontal gas producers that showed similarresults to the H-1 during their flowback periods.They were also picked because they were amongthe highest performers in the field when first puton production. Moreover, feet of net pay open toflow is six times higher in H-2 and three timeshigher in H-3 than in H-1. Additionally, both H-2and H-3 have higher permeability and porositythan H-1.

Nevertheless, the performance comparisonfor the three during the initial flowback periodshows that H-1 and H-2 were the highestproducers with a similar rate of 65 MMcf/d[1.84 million m3/d]. However, Well H-2 achievedthe same rate with a higher flowing wellheadpressure, indicating that it was a betterperformer than H-1. The stimulated Well H-1produced at a higher rate than H-3 with similarflowing wellhead pressures.

Mixing It UpThe mechanical diversion and isolationafforded by this type of system can also be supplemented by chemical diversion. TheKuwait Oil Company (KOC) used a combinationof openhole packers, frac ports and chemicalsto revive a horizontal oil producer in theSabriyah field where production had droppedto zero shortly after the well was completed in2004 (above).14

KOC also chose the StageFRAC servicebecause its mechanical isolation system remainsactive during the life of the well and could beused later to shut off certain zones that wereexpected to eventually experience waterbreakthrough. SXE SuperX concentratedhydrochloric acid [HCl] was used to achievedeep penetration and better etched fractureconductivity. This emulsion fluid is a viscous,highly retarded HCl system designed toovercome acid penetration problems when

14. Al Mutawa M, Al Matar B, Rahman YA, Liu H, Kelkouli Rand Razouqi M: “Application of a Highly Efficient Multistage Stimulation Technique for Horizontal Wells,” paper SPE 112171, presented at the SPE InternationalSymposium and Exhibition on Formation Damage Control,Lafayette, Louisiana, February 13–15, 2008.

> Mauddud well treatment. Based on the reservoir petrophysical model and interpretation, fourintervals in Mauddud C2 and Mauddud D formations (red) in the Sabriyah field were selected forstimulation (inset ). Formation permeability through these zones ranges between 5 and 100 mD acrossthe section. With StageFRAC technology and openhole packer assemblies, the four intervals (red) ofthe 2,562-ft [781-m] openhole horizontal well were compartmentalized into six zones (blue) based onpermeability contrast. The permanent completion resulted in production rates more than five timesthe field average.

8,000

7,000

6,000

5,000

4,000

0 500 1,000

Zone 1

TD 10,520 ft

1,500 2,000

Radial placement, ft

2,500 3,000 3,500 4,000

TVD,

ft

Kickoff point, 6,254 ft MD

7-in. liner point, 8,162 ft MD, 7,601 ft TVD

Zone 5 Zone 4 Zone 2Zone 6 Zone 3

Mauddud C Mauddud C2 Mauddud D

Mauddud B

9,950 to 10,450 ft9,640 to 9,050 ft

9,460 to 9,550 ft9,000 to 9,330 ft


stimulating reservoirs with temperatures higherthan 250°F [121°C]. Standard hydro chloric acidreacts so rapidly at high temperatures that it isimpossible for acid to penetrate, or wormhole,more than a few inches into the formationbefore the acid is rendered ineffective. Deep,live-acid penetration can be achieved only if theacid reaction rate is retarded.

VDA Viscoelastic Diverting Acid was used toensure full coverage across each zone.15 Uponspending, this acid rapidly develops highviscosity in situ and becomes self-diverting. Theviscosity buildup serves as a barrier to reduce thedevelopment of dominating wormholes in theformation and allows movement of the fluids tostimulate other untreated zones. By doing so, itassures treatment throughout the target zone.

Six zones were stimulated within three hours,and the well was flowed back immediately afterthe treatment was complete. The combination oftechnologies allowed successful stimulation of awell that had a 20 to 1 permeability contrastbetween layers using a workover rig instead of amore costly drilling unit. The entire well wasimmediately flowed back and cleaned up to 100%crude oil within two hours. Stabilized

measurement indicates sustained naturalproduction of more than 10,000 bbl/d [1,590 m3/d]of oil—five times the field average and threetimes greater than the best well in the field.

Real-Time ControlChemical diversion is also being used in hydraulicfracturing operations. Through the use ofdiverting fluids that degrade completely after thetreatment, it is possible to stimulate numerouszones in a continuous operation without usingdiverting tools. Recent experience has shown thatthis method of diversion is particularly well-suited to fracture treatment of shale-gasformations. In almost every case, shale-gas wellsmust be hydraulically fractured before they canproduce significant amounts of gas (above). Manyof the new, deeper shale-gas wells are horizontal,and fracturing them can represent a considerableportion of completion costs.

Typically, because of the high cost oftraditional multistage fracturing practices,horizontal shale-gas wells have been limited totwo to four perforation clusters for every 500 ft[152 m] of lateral section. That means that a2,000-ft lateral well, for example, would be

treated in only four stages through 8 to 16 zonesof entry, leaving about 90% of the wellboreuntouched. The optimum approach to shale-gasfracturing would instead be 40 or more smallerstages, clustering the fractures as close togetheras possible.16

When combined with real-time fracturemonitoring, chemical diversion can be used tocontrol fracture propagation as it occurs. Withthe Contact service dynamic category oftreatment systems, engineers use the StimMAPservice to track fracture or refracture creation asthe job is being pumped and then compare theresults with the expected outcome. Then, if thefracture is deviating from its desired course—forinstance, threatening to enter a water zone—engineers can deploy the chemical divertingagent, DivertaMAX effective diversion service formultistage hydraulic fracturing, to redirect it.Slurries containing the DivertaMAX fluid are ablend of degradable materials that cantemporarily block fractures, divert fluid flow andinduce the creation of additional fractures in situand at the wellbore.

This strategy is especially useful in shale-gasplays in which refracturing would seem to be anobvious solution to quickly falling productionprofiles. Perhaps the most active of thesereservoirs, the Barnett Shale in the Fort Worthbasin in north Texas, is a complex reservoir inwhich first-year average production decline ismore than 50%. As a result, many of thesewells—usually horizontal wells with multipletransverse fracture treatments placed acrossthe reservoir—need to be refractured withinfive years of the initial completion. Buteconomics dictate that this must be done moreefficiently than is possible with traditionalmultistage, rig-based fracturing.

16 Oilfield Review

15. For more on viscoelastic acid stimulation: Al-Anzi E, Al-Mutawa M, Al-Habib N, Al-Mumen A, Nasr-El-Din H,Alvarado O, Brady M, Davies S, Fredd C, Fu D, Lungwitz B,Chang F, Huidobro E, Jemmali M, Samuel M and Sandhu D: “Positive Reactions in Carbonate Reservoir Stimulation,” Oilfield Review 15, no. 4 (Winter 2003/2004): 28–45.

16. “Unconventional Gas: Shale Gas,” www.unbridledenergy.com/assets/pdf/About_Shale_Gas.pdf (accessed June 2, 2008).

17. Seale R: “An Efficient Horizontal Open Hole Multi-StageFracturing and Completion System,” paper SPE 108712,presented at the SPE International Oil Conference andExhibition, Veracruz, Mexico, June 27–30, 2007.

18. http://www.slb.com/content/services/solutions/reservoir/unconventional_gas_4.asp (accessed June 12, 2008).

19. http://www.gfz-potsdam.de/portal/-?$part=CmsPart&$event=display&docId=2022464&cP=sec43.quicksearch(accessed June 30, 2008).

> Enhancing production from shale-gas wells. Creating transverse fracturesin horizontal wells greatly increases contact with the shale-gas formation.Transverse fractures (purple) are those in which the direction of the fractureruns perpendicular to the wellbore. They are created by drilling the well inthe direction of minimum horizontal stresses. Longitudinal fractures (green)are parallel to the wellbore and result from fracturing wells drilled in thedirection of maximum horizontal stresses.

TransverseLongitudinal

Vertical stress

Surface

Minimum horizontal stressσh, minMaximum horizontal

stress, σH, max


Summer 2008 17

One operator was faced with this scenario ona Barnett Shale well that initially producedabout 2.2 MMcf/d [62,300 m3/d]. In less than fouryears, production had fallen to less than500 Mcf/d [14,200 m3/d]. Then microseismicmonitoring of the original stimulationtreatments revealed considerable opportunity toincrease formation contact.

The operator employed the DivertaMAXservice in concert with the StimMAP system asan alternative to the prohibitively costlytraditional methods using bridge plugs andpackers on a workstring for isolation. Based onmeasured posttreatment decline rates, theoperator estimates the combination strategy willbe paid out within six months of the stimulation.More importantly, recoverable reserves areexpected to increase by 20% over 20 years.

Another Barnett Shale well was completed in January 2005, and one year later had seenproduction fall from about 2 MMcf/d[56,640 m3/d] to half that amount. Microseismicdata indicated a less than optimal stimulationhad been performed during the third and fourthstages of the well’s original treatment.Production logs run in May 2006 and September2007 also showed that a significant portion of theheel section of the reservoir was not producing,reducing the production rate by half again to500 Mcf/d.

Company engineers decided to perform asingle-stage fracture to stimulate the heelsection of the wellbore. DivertaMAX diversionstages were pumped to allow for movement of thefracture entry point along the lateral. Duringtreatment, diversion plugs were pumped basedon feedback from StimMAP Live monitoring(above). After the treatment, productionincreased immediately from 500 to 1.2 MMcf/d[34,000 m3/d] and payout is expected in ninemonths. The treatment is also estimated to havethe potential to increase recoverable reserves by0.8 Bcf [22 million m3].

Shale Gas: The Next ChallengeSpurred by the low oil prices of the 1980s, the oiland gas industry rapidly developed newtechnology that enabled it to drill longer, moreconvoluted, directional and extended-reachwells. Initially, this effort was aimed atincreasing wellbore contact through naturallyfractured reservoirs that could flow on theirown.17 Today, most of those opportunities havebeen, or are being, exploited, and operators mustlook increasingly to combining the benefits ofextensive formation contact and hydraulicfracturing to attain economic production ratesfrom their horizontal wells.

While that strategy is being applied to manytypes of low-permeability reservoirs, both newand mature, perhaps the most tempting targetfor its application today is in shale-gas reservoirs. Once ignored by operators seeking easier plays and quicker returns oninvestments, these tight-gas formations arecurrently boosting US natural gas reserves. In2007, according to the US Energy InformationAdministration, 3.6 x 1012 m3 [1.3 x1011 Mcf] ofshale gas are technically recoverable from USreservoirs. The challenge is to release it.18

In addition, because of the technology beingdeveloped and proved in the USA, shale-gasreservoirs may soon add significant reservesworldwide. While no commercial shale-gas enterprises are currently known outside ofNorth America, worldwide reserves have been estimated at more than 16,000 Tcf[453 trillion m3] of gas.19

The key to harvesting this potential iscompleting long, high-angle wells efficiently.Technologically, that means placing treatmentsoptimally and accurately in each target zonealong the entire length of wellbore, monitoringand altering the operation in real time and achieving all this at minimum cost and time. —RvF

> Monitoring diversion. Typically, the rapid production decline in the Barnett Shale is attributed to poor stimulation in the heel section (lower right infigures) of the lateral as captured by this StimMAP microseismic monitoring profile (left ). In this case, a later treatment that incorporated DivertaMAXfracture and diversion technology ensured full coverage of all zones. StimMAP Live real-time fracture monitoring of microseismic data indicated that alarge section of the original fracture had been restimulated plus about 25% of new lateral treated (right ).

Dist

ance

, ft

Distance, ft

–1,500 –750 0–3,000 –2,250

–2,250

–1,500

–750

0

Dist

ance

, ft

Distance, ft

–1,500 –750 0–3,000 –2,250

–2,250

–1,500

–750

0


18 Oilfield Review

Managing a Precious Resource

Bill BlackVancouver, British Columbia, Canada

Mohamed DawoudEnvironment Agency – Abu DhabiAbu Dhabi, UAE

Rolf HerrmannAbu Dhabi

Didier LargeauDelft, The Netherlands

Robert MalivaFort Myers, Florida, USA

Bob WillSacramento, California, USA

For help in preparation of this article, thanks to Martin Draeger,Waterloo, Ontario, Canada; Thomas Missimer, Fort Myers,Florida; and Dominique Pajot, Paris.

AIT (Array Induction Imager Tool), AquaChem, AquiferTest Pro,CMR (Combinable Magnetic Resonance tool), CTD-Diver,Diver, Diver-NETZ, DSI (Dipole Shear Sonic Imager), ECLIPSE,ECS (Elemental Capture Spectroscopy sonde), FMI (FullboreFormation MicroImager), Hydro GeoAnalyst, MDT (ModularFormation Dynamics Tester), Petrel, Platform Express, RFT (Repeat Formation Tester), RST (Reservoir SaturationTool) and Westbay are marks of Schlumberger.

Many parts of the world experience seasonal or long-term imbalances between

freshwater supply and demand. Groundwater resources are increasingly called upon

to offset a greater share of these shortfalls. To help manage and sustain dependable

groundwater supplies, resource managers are turning to advanced geophysical

logging technology—much of it pioneered in the oil field—in combination with

innovative downhole monitoring and sampling techniques.

Saline ocean water97.4%

Surface water0.3%Other sources0.9%Groundwater30.1%Ice caps andglaciers68.7%

Fresh water2.6%

21502schD5R1:Format Layout V7.2 8/25/08 6:41 AM Page 18

Summer 2008 19

Given the balance of water resources on ourplanet, it is amazing that we have ever hadenough to drink. Although Earth is known as “the water planet,” most of that water isundrinkable. Covering some 70% of the Earth’ssurface, about 97.4% of the water is salt water,leaving only about 2.6% as fresh water. Most ofthat fresh water is largely inaccessible, since it isfrozen in ice caps and glaciers; much of theremaining portion lies below ground.

In most cases, water does not come to us—either we go to it, or we devise ways of delivering itto our faucets or taps. Most of the world’spopulation—about 60% and growing—lives nearcoastal areas, while many who live inland tend tobe near surface water sources, such as streams,lakes and springs. With population growth andassociated increases in domestic water consump -tion, industrial use and agricultural demand, thewater resources at these locations are increasinglysubject to depletion, environmental degradationand pollution. Demographics, infrastructure,pollution and drought set the stage for a broadrange of problems that ultimately affect theavailability and distribution of one of our mostvaluable resources.

An obvious solution is to bank thisresource—storing water when it is abundant, foruse when it is not. This solution brings animportant issue to the forefront: namely, how tostore the water for maximum recovery. Surfacereservoirs, water-retention ponds and lakes alllose water to evaporation, or can be renderedundrinkable through contamination by pollu -tants. In addition, the water stored in tanks candeteriorate considerably over time because ofbacterial growth. And although surface storagetanks are an alternative, they tend to beexpensive, are subject to eventual structuralfailure, and are vulnerable to natural or man-made calamities.

Yet another approach is to deposit excesswater into existing groundwater aquifers duringthe wet season (typically winter months, whensurface-water supply exceeds demand), and thenwithdraw it during dry spells. This approach,known as aquifer storage and recovery (ASR) orartificial recharge (AR), is in some waysanalogous to underground gas storage; in thiscase, a known reservoir (or aquifer) is filled withfresh water instead of natural gas.1 Capacity is adistinct advantage of aquifer storage over surfacestorage tanks; for some aquifers, this capacitycan range to billions of gallons of water. Aquifersystems can accommodate multiyear storage

below ground, protecting the water supply fromevaporation while maintaining good waterquality. Above ground, water wells can often belocated where they are most needed, and becausethe surface installation occupies little space, thecosts of land acquisition can be minimized. Bystoring large volumes of water below ground, theneed for constructing and maintaining largesurface reservoirs is reduced.

Underground reservoirs can be charged, orlater recharged, through downhole waterinjection. In a process similar to that used in theoil field, the injection process requires awellbore to place water into a subsurfacereservoir. If the aquifer is unconfined, having nolow-permeability barriers, or aquitards, whichcould impede the flow of water between theground surface and free-water table, then theaquifer can also be charged through downwardpercolation of surface water. The percolationprocess involves spreading water across theground in shallow ponds, allowing it to infiltrateinto the underlying water table. In other cases,river flow is altered to prolong the exposurebetween river and riverbed, thus encouraging thewater to percolate into the ground rather thanflow to its eventual meeting with the ocean.

The water that is injected or spread onto theground needs to come from somewhere.

Naturally occurring fresh water would be best,but usually there isn’t enough to sustain thiseffort throughout the year. For instance, someareas in low-latitude equatorial regions have awet winter season—often charac terized by an overabundance of water—that does notcompensate for the much drier summer season.In slightly higher latitudes, subtropical zones,such as those of the Middle East, may be so aridthat rain barely enters into the water-useequation. These desert areas may experience onlya fraction of an inch of rainfall each year. Otherareas, such as southern California, may haveaccess to water imported through aqueducts, butare motivated to minimize consumption by thehigh cost of this imported water.

In each case, water resource managerssupplement the water stored in aquifers with“new” or repurposed water (above). In theMiddle East and other locations, this new wateris obtained through desalination, as waterauthorities remove salt from seawater or frombrackish groundwater produced by water wells.To supplement imported water conveyed byaqueduct to southern California, water resourcemanagers treat waste water, filtering and

1. For more on underground gas storage: Brown K,Chandler KW, Hopper JM, Thronson L, Hawkins J, Manai T, Onderka V, Wallbrecht J and Zangl G:“Intelligent Well Technology in Underground GasStorage,” Oilfield Review 20, no. 1 (Spring 2008): 4–17.

> Freshwater sources. Although many people think that our freshwatersupplies come from rainfall, snow melt or underground springs, waterresource management often requires a blend of sources to reclaim watersfor domestic, agricultural and industrial use.

User

Groundwater

Treatedwaste water

Desalinatedwater

Surfaceflows

Sources of Water

Surfacestorage

Undergroundstorage


sanitizing it to meet drinking-water standardsbefore pumping it into the aquifer. Each situationrequires a targeted approach to the water usedfor recharging aquifers.

Injected or percolated water will affect thenative groundwater contained within the aquifer.The newly introduced water tends to displace thenative water contained in the aquifer, creating adistinctive bubble of fresh water surrounded by aprotective buffer zone of mixed waters (left).This buffer exhibits a gradational densitycontrast between the native and injected waterscaused by differences in chemical formu lationand dissolved solids in each body of water.

As with gas storage, subsurface reservoirsmust be thoroughly understood in terms ofheterogeneity, boundary conditions, geochemistry,hydraulic properties and water quality. This aim isachieved through hydrogeologic studies, surfacegeophysics, drilling and borehole geophysics,aquifer performance testing, and detailed in-situmonitoring and sampling of the water over time.Groundwater modeling and numerical simulationhelp delineate the aquifer spatially, characterizeits inflow (recharge) and outflow (discharge)rates, and forecast the system response to various stresses.

Subsurface reservoirs, produced fluids, water -floods, well logs, fluid samples and computermodels are concepts that are not exclusive to thewater industry. There are many obvious simi -larities between engineering, geophysical andgeological concepts that pertain equally to oil andgas production as well as to water production.Indeed, the history of oil and gas production isreplete with fields that produce more water than hydrocarbons.2

With experience in measuring and charac -terizing subsurface reservoirs and fluids for theoil and gas industry, Schlumberger is alsofocusing attention on a different kind ofreservoir, with the aim of helping resourcemanagers better evaluate, monitor and care forgroundwater resources. Through investments inpersonnel and technology, the company has builta portfolio of expertise in aquifer evaluation,characterization, monitoring and groundwaterresource management (left). Over the years,Schlumberger has completed projects to helpmeet future water demands on six continents.Schlumberger engineers, hydrogeologists andgeophysicists routinely face challengespertaining to ASR, water resource management,coastal-zone groundwater management, ground-water monitoring and water solutions for minesand sensitive environmental sites.

20 Oilfield Review

> Buffer zone. A bubble of injected fresh water is surrounded by a bufferzone where differences in density and total dissolved solids between theinjected and native waters tend to blend together.

Typical depth 60 to 900 m

Wellhead

Bufferzone Injected fresh waterNative

water

> Acquisition history. At the turn of the century, Schlumberger began aprogram to expand its expertise and capabilities in the field of water services.

2000

Year

Westbay: developed multilevel completions, along with pressure and sampling probes for detailed characterization and monitoring of deep, complex aquifers

Company Acquired

2001 Van Essen Instruments: developed maintenance-free data loggers for groundwater monitoring networks

2001 Saracino-Kirby-Snow: provided consulting services specializing in conjunctive use, groundwater banking and groundwater quality protection

2005 Waterloo Hydrogeologic: developed environmental software focused on data management, data analysis and groundwater modeling and visualization

2007 Missimer Groundwater Science: provided consulting services with expertise in well field design, groundwater modeling, water resource evaluation and aquifer storage and recovery

2007 Water Management Consultants: provided consulting services specializing in environmental baseline reports, impact studies and mine dewatering, with staff expertise in hydrogeology, civil engineering, hydraulics and geochemistry


Summer 2008 21

This article describes some of the issuesaffecting our quest for drinkable water and thetechnologies used to maintain our watersupplies. Case studies from the UAE and the USAwill shed light on some of the unique problemsfacing water authorities, and the technologiesthey employ in managing and preserving thisprecious resource.

Aquifer AssessmentAquifer quality varies greatly from one reservoirto another. It is influenced both by geology—which can change from shallow gravels and sandsto deep fractured rocks and karsts—and by waterquality, which ranges from fresh to brackish tosaline.3 To understand and predict the behavior ofunderground storage and withdrawal systems,numerous aquifer characteristics must beaccurately measured and delineated.

Aquifer assessment requires a multidimen -sional evaluation of local and regional geology,hydrogeology and hydrology. Both the rockthrough which the water travels and the rock inwhich it will be stored affect the properties of thenative groundwater. Hydrologists need tothoroughly characterize the basin surroundingthe aquifer and the regional tectonics thatshaped the basin to understand storage and flowmechanisms. In addition to analyzing thechemical composition of native groundwater,they must analyze imported water and screen itfor contaminants.

Aquifer managers must ensure that anyimported water meets certain quality standardsfor the concentration, source and type of con -taminant. They need to screen for commonpollutants, such as nitrogen from fertilizerrunoff, in addition to biological contami nantsfrom farm waste, stormwater runoff or sewage.The chemical composition of imported watermust be tested for compatibility with the nativegroundwater and its aquifer; the conse quences ofincompatibility may range from mineraldeposition to clogging or other such degradationof the aquifer (above right). On the other hand,some reactions may actually improve the qualityof stored water by decreasing the concentrationsof organic compounds and microorganisms. Thisimprovement in water quality is considered anatural by-product of subsurface water storage.

Water samples alone do not provide sufficientdetail for assessing the viability of aquifer-storageschemes, so high-resolution borehole and surfacegeophysics plays an increasingly important role in

aquifer characterization. The similarity betweenevaluation technologies for oil fields andgroundwater well fields grows as reservoirmeasurement and assessment techniquesdeveloped for the oil and gas industry are beingadapted for evaluation of groundwater resources.4

From the earliest days of wireline logging, thesearch for oil and gas has been intimately tied tothe detection of water. Indeed, the foundations ofoilfield petrophysics were built largely oncalculations of water saturation (Sw), based onwireline log measurements of spontaneouspotential and formation resistivity. Originally,oilfield petro physicists would calculate Sw todetermine the percentage of a rock’s pore spacethat was saturated by water. From thispercentage, they could infer that the remainderof the pore space was filled by hydrocarbons.

Earlier, a measurement-technology gapexisted between the oil industry and the waterindustry. The technology gap was driven in largepart by the relatively low market value placed on

water: so long as water was cheap and plentiful,there was no economic justification for applyinghigh-tech approaches to development ofgroundwater resources. This gap is progressivelyclosing as easily accessible freshwater resourcesgrow scarce, and alternative water supplies areneeded to meet growing demands. A basicborehole geophysical logging program for a wellfield project might call only for a caliper, gammaray, resistivity, spinner flowmeter and perhaps asonic log—a common suite of logs for mostwater-well logging companies. For detailedaquifer characterization, a more advanced suiteof logs is required. Schlumberger logging toolsonce destined exclusively for the oil patch areincreasingly used for evaluating groundwaterreservoir heterogeneity, water quality andaquifer hydraulics.

ASR feasibility studies utilize a broadassortment of data, much of which can beobtained through surface and boreholegeophysical measurements, with the rest

2. Arnold R, Burnett DB, Elphick J, Feeley TJ III, Galbrun M,Hightower M, Jiang Z, Khan M, Lavery M, Luffey F andVerbeek P: “Managing Water—From Waste toResource,” Oilfield Review 16, no. 2 (Summer 2004): 26–41.

3. A karst is a type of topography formed through dissolutionof carbonate rocks. Sink holes, caves and pock-markedsurfaces are typical features of karst topographies.

4. Well field, a common term for groundwater professionals,is used to describe an area in which more than one welland associated piping have been installed for the purposeof injecting, withdrawing or monitoring groundwater.

> Potential fluid interactions in ASR systems. Groundwater managers mustensure that the fluid chemistry of injected water is compatible with that ofthe aquifer. Incompatibility can generate unwanted effects and reduce therecovery efficiency of the overall system.

Leaching of trace elements, such as arsenic, molybdenum, nickel and uranium, incorporated into or sorbed onto destabilized sulfide minerals

Clogging due to algae growth

Carbonate disequilibrium

Cause Effect

Introduction of dissolved oxygen

Cation exchange

Oxidation of organic matter

Decrease in salinity

Biological activity

Scaling and associated reduction in permeability

Clogging due to iron oxyhydroxide precipitation

Decrease in hardness and increase in sodium concentration of stored water

Dissolution of carbonate minerals and increase in permeability

Clay swelling and dispersion

Diluted concentrations of dissolved organic carbon and some disinfection by-products

Biologically mediatedredox reactions

Increase in concentrations of reduced iron and manganese


obtained through drilling or actual sampling (left). Beyond characterizing basic formationproperties, aquifer managers needgeomechanical, structural and geochemical datato investigate an aquifer. From a basic frameworkbuilt on lithology, stratigraphy and structure,groundwater-flow and solute-transport modelsare developed, then populated with aquifersystem details obtained from a variety of sources.Pore-size distribution, total and effectiveporosity, bulk density, mineralogy, bed thicknessand rock mechanical properties can bedetermined from well log data (next page,bottom left). Understanding the orientation andpermeability of fractures and faults, which oftenserve as conduits for fluid flow, is critical for localand regional geomechanical assessments, and fordesigning hydrologic tests and groundwatermonitoring networks. When combined withcrosswell and surface seismic surveys, boreholedata, as provided by an FMI Fullbore FormationMicroImager, are useful in determining the type,geometry and orientation of fractures.

In-situ stress can be estimated frommeasurements of borehole breakout, induceddrilling fractures and minifrac testing. Othergeomechanical properties can be calculatedfrom acoustic and density log data. Theseproperties are used for assessing and predictingregional and local stresses pertinent to wellboreand completion design, and for predictingsubsidence or deformation caused by ground -water pumping or injection.

Hydrogeologic properties, such as hydraulicconductivity, transmissivity and storativity,vertical flow rate, capillary pressure and porepressure, reveal a great deal about howgroundwater will flow through an aquifersystem.5 Just as important, these properties helpaquifer managers characterize contaminanttransport. The quality of the groundwater isaffected by elemental concentrations, mineralogy,lithology and salinity. To measure theseconcentrations, geochemical logging tools suchas the ECS Elemental Capture Spectroscopysonde are used for estimating groundwatersalinity and measuring the abundance of rock-forming elements such as silicon, calcium, iron,sulfur, carbon, oxygen, titanium and aluminum.

Aquifer assessment is a continual process,and once a groundwater reservoir is utilized foraquifer storage, resource managers must bevigilant in monitoring and identifying conditionsthat might compromise water quality.

22 Oilfield Review

> Aquifer log. A section through an unconfined aquifer typical of thoseseen in the UAE shows the top of a free-water table (blue line) beneath asandy, hydraulically conductive unsaturated zone. This aquifer has beendivided into three zones, as demarcated in the mud log lithology column(Track 3). Zone 1, in the upper part of the section, exhibits the bestcombination of lithology, porosity and permeability (Tracks 3, 4 and 5,respectively). Zone 2 shows a decrease in these properties. Zone 3, atransition zone above the clay base, exhibits some porosity and lowpermeability; it represents a distinct facies change at the bottom of theaquifer. Beneath Zone 3, an erosional contact, or unconformity, marks thetop of a layer of clays and shales (green line). These low-permeabilitysediments form an aquitard that impedes groundwater flow to adjacentpermeable beds.

35

40

45

50

55

60

65

70

75

80

85

90

30

25

Dive

r 1Di

ver 2

PVC

casi

ngPV

C sc

reen

PVC

casi

ng

Zone

1Zo

ne 2

Sand

Zone

3Sh

ales

and

clay

No

kM

ed k

>1,

500

mD

Low k > 500 mD

Low

k >

500

mD

No

kM

ed k

>1,

500

mD

PermeabilityPermeability

0 10,0000 40Porosity

Mud

Log

Litho

logy

Data

Log

ger

Posi

tion

Com

plet

ion

Mea

sure

dDe

pth,

m

Low

-per

mea

bilit

yaq

uita

rdSa

tura

ted

aqui

fer

Unco

nfin

ed zo

ne

Base of aquifer

Top of aquifer

Interpretation

% mD


Summer 2008 23

Groundwater MonitoringPractices for monitoring groundwater reservoirsdiffer substantially from those of oil and gasreservoirs. Groundwater monitoring relies onnetworks of nonproducing observation wells,drilled and dedicated solely to measuringconditions within the hydrogeologic system thatmight portend changes in groundwater quality orquantity. The relatively shallow depths of mostaquifers make the drilling of such wells arelatively inexpensive proposition, and the criticalneed for monitoring to safeguard water suppliesmakes these wells an absolute necessity. Althoughgroundwater production wells are also monitored,the observation wells provide access to pointsdistributed laterally and vertically throughout thegroundwater basin, providing a three-dimensionalnetwork that allows resource managers to survey awide sweep of the overall system.

When placed on the flanks of one or moreproducing wells, a network of observation wellscan detect changes in the groundwater system towarn resource managers of impending threats tothe aquifer. Groundwater monitoring networkscan provide baseline data for mapping the spatialand temporal distribution of various watercharacteristics and help identify short-termchanges to groundwater flow caused by pumping,natural recharge and discharge, and agriculturaland industrial use. Monitor-well data also provide

a real-time accounting of water use, which can behelpful for ensuring user compliance withregulatory guidelines. Finally, groundwater moni -toring is valuable for calibrating the computermodels of the hydrogeologic system.

A basic monitoring well typically consists of ashallow wellbore, drilled slightly beyond thedepth required to monitor a specific interval ofthe aquifer, and a standpipe or casing with aslotted screen (right). The 1- to 4-in. [2.5- to 10-cm]diameter standpipe or casing is lowered into thehole so that the slotted-screen section lies acrossthe interval to be monitored. A filter pack ofrounded gravel or graded sand is backfilledaround the screen, then the remainder of thehole is sealed with cement or bentonite grout.Groundwater enters the cased well through thescreen and rises inside the casing until theheight of the water column balances thegroundwater pressure in the monitored interval.

Water quality can be determined byoccasionally pumping water from the well andanalyzing its chemistry in a laboratory. Aquiferfluid pressure is monitored by measuring theheight of the water column in the well. The waterlevel can be obtained by simply lowering ameasuring tape into the observation well andrecording the distance from the top of the well tothe water level in the cased hole. Until recently, groundwater measurement

data were obtained manually. This practice wastime-consuming, requiring a trained field techni -cian to visit each site in a well field that couldspan more than 100 km2 [39 mi2]. Data gatheringwould be further complicated in areas whereaccess to obser vation or monitoring wells wasrestricted by heavy snowfall, flooding oruncooperative landowners. Depending on theapplication, measurements might be obtainedonce a week in the best scenarios, or even onceper season—or less—leading to fewer points forestablishing trends and ultimately diluting thevalue of the data.

The process of gathering data frommonitoring wells has grown much easier assensor technology has evolved. One approach totracking water-level changes in monitoring wellsinvolves installation of a pressure transducer atsome depth below the surface of the watercolumn, with pressures measured by thetransducer used to calculate water level. In thepast, these pressure sensors were hard-wired to

5. Storativity is a depth-integrated percentage of the amountof groundwater that can be released from storage due todepressurization of a confined aquifer; it is defined asthe volume released from storage per unit decline inhydraulic head, per unit area of the aquifer. Transmissivityis the hydraulic conductivity divided by the verticalthickness of the aquifer. These properties both measurean aquifer’s capability to discharge groundwater.

> Advanced borehole geophysical logging suite. Logging tools oncereserved for oil and gas applications are increasingly called on for precisecharacterization of groundwater systems.

Used to determine strike and to evaluate dip of bedding and fractures, and for evaluating rock or sediment texture

AITArray InductionImager Tool

Obtains formation electrical resistivity across five highly focused depths of investigation (from 10 to 90 inches) to delineate bedding, evaluate drilling fluid invasion and estimate formation water salinity

DSIDipole ShearSonic Imager

FMIFullbore FormationMicrolmager

Produces fully oriented 3D images of the electrical resistivity or acoustic properties of the formation surrounding a borehole

RSTReservoir SaturationTool

Measures spectral gamma ray activity induced by electronically generated neutrons to determine elemental weight fractions of a number of key rock-forming elements

CMRCombinable MagneticResonance tool

Measures the magnetic resonance response of free hydrogen in a formation’s pore space and is used to evaluate pore-size distribution, effective porosity, total porosity and hydraulic conductivity

MDTModular FormationDynamics Tester

Collects fluid samples, measures aquifer pressures, obtains permeability and anisotropy data and provides minihydrofrac tests

Measures interval transit times of compressional and shear waves, along with Stoneley slowness, to determine stress anisotropies, geomechanical properties, porosity, permeability and fracture permeability, and to calibrate surface seismic data

Tool Function

> Typical monitoring well design. Here, a 4-in. [10-cm] polyvinyl chloride (PVC) casing and PVCscreen are installed in an 8½-in. [22-cm] borehole.

Gravel pack

4-in. screen

Water level

Top of clay

Total depth

Grout

4-in. casing

21502schD5R1.qxp:21502schD5R1.pdf 10/6/08 9:41 PM Page 23

the surface to enable pressures to be displayed orrecorded. However, the surface equipmentneeded to be powered and could be targeted for vandalism.

By the 1990s, downhole electronics hadadvanced to the point that a small, self-containedunit incorporating a pressure sensor, tempera turesensor, data-storage memory and power supplycould be permanently deployed in a monitoringwell. The first such sensor package, known as aDiver automatic groundwater logger, wasdeveloped in The Netherlands by Van EssenInstruments. The Diver family of instruments forlong-term monitoring of groundwater and surfacewater parameters has proved to be reliable andeasy to install (below). Diver data loggers can beprogrammed to record pressure and temperaturemeasurements at any specified time—in starkcontrast to data obtained by hand.

Data loggers are becoming important formanaging coastal aquifers, where they are usedto monitor salinity over time. In these aquifers,the interface between fresh water and seawaterexists in a state of dynamic equilibrium, movingback and forth with increases or decreases ininland pumping of the water table, or withvariations in tides, storms or atmosphericpressure. This movement is relatively slow, and aconsiderable span of time may pass before risingsalinity is noticed in area water wells. A densemonitoring network is usually required to detectsaltwater encroachment, and this applicationrequires a different kind of data logger. The CTD-Diver automated data logger measureselectrical conductivity (EC), in addition totemperature and pressure. When correlated tolaboratory measurements of total dissolvedsolids, the EC readings can be used to tracksalinity changes over time.

The Diver family of data loggers can operateunattended for up to 10 years. These data loggersprovide a cost-effective means of trackingchanges in monitoring wells, but the data stillneed to be collected in the field. In the mostbasic operation, the data logger is suspended in awell on a steel cable. The unit must be recoveredfrom the well to download data from memory.

Advances in electronics and data communi -cation continue to open new modes forcollecting groundwater data. One approach usesa Diver data cable, which permits directreadout, or access to data logger memory fromthe wellhead without having to retrieve thesensor from the wellbore. However, a greaterdegree of flexibility in collecting and managingwell field data can be obtained with the Diver-NETZ data management system forgroundwater monitoring networks (above). Withits wireless connection range up to 150 m [490 ft],

24 Oilfield Review

> CTD-Diver data logger. This compact sensor, 183 mm [7.2 in] long, is suspended inside amonitoring well to record pressure, temperatureand electrical conductivity of the water.

by

Van

Ess

en In

stru

men

tsA

Schlu

mberg

er pr

oduc

t

ww

w.v

anes

sen.

com

R

> Wireless data collection. The Diver-NETZ system enables monitoring technicians to rapidly collectaccurate groundwater data without the need to visit every wellhead.

Traditional data collection: visit each well

Diver-NETZ data collection:receive data from multiple wells

Diverdata logger

Divertransmitter

Hand-heldreceiver


Summer 2008 25

field technicians are able to locate and collectdata without having to walk to each well.Avoiding site obstacles, fenced property, steepinclines and river crossings increases safetywhile boosting data-collection rates.

In 2007, Schlumberger conducted a Diver-NETZpilot study in cooperation with the city of Guelphin Ontario, Canada. The test area covered 16wells scattered across 12 facilities, and extendedmore than 150 km2 [58 mi2]. Before the study,the city’s monitoring technicians would manuallycollect groundwater data by locating each well,unlocking its cap, lowering a water-level meterand taking readings. The Diver-NETZ systemproved to be 70% more efficient than the manualmethod. In turn, a two-person, 75-hour collectionexcursion was reduced to a one-person, 4-hourtrip. The new data-collection routine resulted ina dramatic decrease in cost—from 10 Canadiandollars, down to 25 Canadian cents per datapoint. Meanwhile, the amount of data collectedincreased from 300 to 3,000 data points a month.This approach was instru mental in meeting city-legislated compliance efforts while stayingwithin budget.

Although data loggers have proved to besimple and cost-effective for monitoring ground -water, a different approach is required for testingindividual zones within a well. For the past30 years, geoscientists seeking high-resolutiongroundwater monitoring capabilities have usedsystems that enable monitoring of groundwaterconditions at multiple points in a borehole.6 Oneof the first of these technologies to becommercialized was developed in Canada byWestbay Instruments of Vancouver, BritishColumbia—a company later acquired bySchlumberger. Originally designed for moni -toring pore pressures in landslide-prone areas,the technology spread fairly quickly to othergeotechnical and construction applications, such as site characterization for geologicrepositories, investigations of contaminatedgroundwater and, eventually, general ground -water resource management.

The Westbay multilevel groundwater charac -terization and monitoring system was inspired, inpart, by oilfield technology. Groundwatergeoscientists had long known that even slightchanges in pressure can affect the chemistry offluid samples obtained downhole. In addition,they wanted to prevent fluid samples in onesection of a well from mixing with those in

another level. Recognizing that these sameproblems had been addressed by oilfield wirelinetools such as the RFT Repeat Formation Tester,Westbay system designers adopted a similarapproach for groundwater sampling. Thismonitoring system is designed to collectsubsurface samples and measurements atmultiple discrete positions within a single well.

The Westbay system uses modular casing,packers, port couplings and associatedcomponents to preserve wellbore integrity whileproviding the capability to isolate numerouszones for testing and monitoring.7 The Westbaycasing is sealed along its entire length,preventing groundwater from flowing up or downinside the casing, while inflatable packers on theoutside of the casing seal the borehole intomultiple testing or monitoring zones, preventingflow up or down the annulus (right). Specializedwireline probes run within the Westbay casingcan locate and operate valved ports to accessborehole fluids in the zones between packers.

These probes fulfill a variety of roles, such asinflating and deflating packers, measuringgroundwater temperature and pressure,collecting in-situ samples and executinghydraulic or other system-specific tests. A specialsampling probe provides the capability to collectand preserve discrete fluid samples at formationpressure, with minimal disturbance and withoutrepeated purging. The Westbay casing assembliesgenerally consist of polyvinyl chloride (PVC) orpolyurethane plastics for most groundwaterenvironments, but can also be made of steel towithstand higher pressures and temperatures.

The groundwater industry is graduallyacknowledging the value of vertical profiles ofpressure and chemistry based on substantiallymore measurements—30 to 40 monitoring pointsthrough a 120-m [390-ft] interval—than washistorically considered the norm. This kind ofhigh-resolution, long-term data plays animportant role in advanced suites of modelingand interpretation programs used to managegroundwater resources.

Getting the PictureTo track changing conditions within an aquifer,groundwater managers require not only up-to-the-minute monitoring data, but also static and dynamic models of the aquifer itself.Understanding the risks implicit in such aquiferchanges requires an in-depth understanding of

the aquifer, its hydraulic conductivity, storativity,effective porosity, groundwater flow directionand other aspects of its hydrogeologic regime. Togain insight into their reservoirs, groundwatermanagers evaluate the monitor-well and water-sample data with the aid of models andsimulations. These programs help hydro -geologists define the groundwater inflow, outflow

6. Ellis D, Engelman B, Fruchter J, Shipp B, Jensen R, Lewis R, Scott H and Trent S: “EnvironmentalApplications of Oilfield Technology,” Oilfield Review 8,no. 3 (Autumn 1996): 44–57.

7. One company installed 60 inflatable packers in a well;however, the length of the casing string is the only realconstraint on the number of packers that the Westbaysystem can accommodate. To date, the maximum depthof installation was set at 1,270 m [4,167 ft]—considereddeep by most groundwater standards.

> Multilevel monitoring technology. The Westbaycompletion system allows monitoring of fluidtemperature and pressure, collection of fluidsamples, and injection or withdrawal of fluids totest hydraulic conductivity from multiple zoneswithin a single borehole.

Measurementport

Pumpingport

Packer

Samplerprobe

Samplercontainer

Sealedconnections


and recharge characteristics of the aquifer. Aswith other tools described previously, some typesof reservoir-analysis software for oil and gasfields have been adapted to model and simulatethe behavior of groundwater systems.

The myriad processes and analyses requiredto evaluate an aquifer are best managed througha systematic workflow.8 Schlumberger water-service specialists have developed workflows thatlink existing groundwater software with oil andgas software applications. The Hydro GeoAnalystproject workflow helps resource teams constructand evaluate hydrologic models and simulationsin near-real time. This workflow lets hydro -geologists, hydrochemists, geologists and watermanagers share data and results within the sameenvironment, driving collaboration betweendisciplines. In this workflow, logs and other field-collected data are combined with models to verifycorrelations and validate interpretations.Interactive 3D views and cross-sectional viewersenable resource managers to visualize complexrelationships between geology and hydrogeology,while 3D animations of groundwater flowsprovide additional insight and a means toexamine various production scenarios (above).

Using Petrel seismic-to-simulation modelingsystems, aquifer managers can import well logs,surface resistivity (time-domain electromagnetic)data and surface seismic data to build detailed 3Daquifer models. They can further condition theaquifer model with input from 3D surfacegeophysical surveys, geological maps, faciesmodels, seismic attributes and geostatisticalmodels. To understand flow and mass transport inhighly complex conditions, they can import thePetrel aquifer model into ECLIPSE reservoirsimulation software. The ECLIPSE programprovides well-performance tools not previouslyavailable to the water industry and can accom -modate special conditions, such as incorporatingdensity effects for modeling the behavior of salineaquifers, saltwater intrusion into coastal aquifers,or unconfined conditions of free-water-tableaquifers, rainfall and surface river systems. It can also model nonaqueous phase liquids such asoil or gasoline contaminants, along with thechemical reactions that ensue as they movethrough the system. In addition, the program cansimulate physical features such as directionalpermeability, vertical equilibrium, dual porosityand dual permeability.

The following case studies highlight theapplication of oilfield and groundwater-specific technologies.

Aquifer Storage and Recovery in Arid EnvironmentsClimate significantly influences a country’s water supply. Arid climates in particular—charac terized by low precipitation and highevaporation rates—can hamper the domestic,agricultural and industrial endeavors of a country.In the United Arab Emirates, Abu Dhabi has longdealt with challenges imposed by climate, butnow is facing increased rates of water consump -tion to meet the needs of a burgeoning populationand economy.

Deterioration of nonrenewable aquifers oraquifers that are slow to recharge has promptedthe Abu Dhabi Emirate and most countries of theGulf Cooperation Council (GCC) to rely ondesalination as a primary source of domesticwater. It has been argued that the best long-termsolution for ensuring regional GCC watersupplies would be a network of large-scale desali -nation plants. One problem facing Abu Dhabi,however, is the vulnerability of desalinationplants to pollution or other calamities that couldforce emergency shutdowns. The concern is thatwater supplies would not be sufficient to meet

26 Oilfield Review

8. Herrmann R, Pearce M, Burgess K and Priestley A:“Integrated Aquifer Characterization and NumericalSimulation for Aquifer Recharge and Storage at MarcoLakes, Florida,” in Hydrology: Science & Practice for the21st Century, Proceedings of the British HydrologicalSociety International Conference, Vol. 1. London: BritishHydrological Society (2004): 276–283.

9. Eutrophication is a process in which runoff from animalwaste, fertilizers or sewage increases the nutrientdensity of a body of water, producing an overabundanceof plant life. Subsequent decomposition of these plantsdepletes oxygen in the water, leading to the death ofaquatic life.

> Groundwater project workflow. The Hydro GeoAnalyst workflow approach links existing groundwater characterizationapplications with those developed for the oil and gas industry.

Data Gatheringand Input• Information management• Geographic information system database

Surface Imaging• Mapping• Electromagnetic survey interpretation

Geophysical Log Interpretation• Well correlation• Formation properties

Data Analysis• Facies modeling• Fault modeling• Fracture modeling• Hydrodynamic test analysis

Uncertainty Analysis• Upscaling processes• Aquifer property populationConceptual Model

• 3D geological model• Hydrogeologic conceptual model

3D Flow and Mass TransportSimulation• Saturated and variably saturated conditions• Density-dependent modeling• Geomechanics

Calibration• History-matching• Postprocessing presentation


Summer 2008 27

demand during prolonged shutdowns. Also, whilewater production from desalination plants isfairly constant, the demand for water hassignificantly increased, owing to rapid growth invarious development sectors and increases inpopulation. To provide a strategic water reservefor emergency conditions, water managersproposed an ASR project to develop a 30-billion imperial gallon [136,382,756 m3]freshwater reserve.

Desalination and ASR technologies go hand-in-hand in this arid region. More than 90% of thefresh water supplied for domestic use in AbuDhabi is produced through desalination ofseawater or brackish groundwater. Excessproduction from desalination plants can bestored for future demand, and certain shallowaquifers in the area may be well-suited for this purpose.

In arid climates, water storage in naturalaquifers provides a favorable alternative tostorage in surface reservoirs, lakes or tanks (above). Compared with surface storagemethods, ASR projects enjoy significantlyreduced construction costs along with limitedenvironmental effects, such as evaporative lossof surface waters, eutrophi cation and thepotential for catastrophic failure of dams andtheir reservoirs.9 The minimal surface footprint

> ASR versus surface storage. A comparison between surface storage in tanks and subsurface storage in anaquifer for potable water reveals distinct advantages of storing water underground. To store 4 billion imperialgallons in an aquifer, an equivalent of 200 surface tanks with a capacity of 20 million imperial gallons would benecessary. The footprint of these surface tanks (black circles) may encompass an area 1.5 km by 3 km [0.9 by 1.9 mi],with considerable negative environmental impact. The freshwater bubble for subsurface storage (red oval)equivalent to the storage capacity of 200 surface tanks induces pressure changes during injection or recoveryover a larger area (green oval).

20 millionimperial gallons


of ASR projects belies the extent of theirsubsurface storage capacity, which may extendto billions of gallons (right).

This approach can help Abu Dhabi watermanagers reach their goal of saving a 180-daysupply of fresh water for the main cities of Al Ainand Abu Dhabi. This strategic target wouldsafeguard against unforeseen problems thatmight curtail production from area desalinationplants.10 Their plan called for aquifers in theeastern region of the Abu Dhabi Emirate to beartificially recharged with fresh water from adesalination plant on the Indian Ocean coast atQidfa, in the Emirate of Fujairah (below). Thedesalinated water will be injected into theaquifer system for about 200 days per year, afterwhich the groundwater stores can be used to

supplement potable water supplies duringsummer months and other peak-demand periods.During water emergencies or to supplement peakdemand, the final ASR well field will operate at aproduction capacity of 20 million imperialgallons per day [90,922 m3/d], with potentialexpansion to 100 million imperial gallons per day[454,609 m3/d].

A study initiated by the Abu Dhabi govern mentand managed by the Environment Agency–Abu Dhabi (EAD) was conducted bySchlumberger to test the potential of an ASRsystem. The initial phase of the project entailedlocating a potential ASR site. This phase called for a review of geological and hydrogeologicinformation: lithology, geophysical logs, water-level records, water-quality data, seismic data and

28 Oilfield Review

> ASR surface footprint. An injection site in AbuDhabi reveals little about the asset stored belowground. This dual-purpose installation allows for injection (right-hand side) and recoveryassisted by a submersible pump (larger pipe onthe left-hand side). This wellhead is one of twoused during ASR pilot testing, and just one wellwas required to inject about 30 million imperialgallons [136,383 m3], or nearly 1½ times thevolume of a large surface tank.

> Shwaib ASR site. Fresh water from the desalination plant in Qidfa provides injection water for an ASR site at Shwaib, north of the oasis city of Al Ain.From Qidfa, the fresh water is pumped 170 km [105 mi] through a dual pipeline (blue line), over an 800-m [2,625-ft] mountain range on its way to the ASRwell field. The ASR site is located along the western side of an Omani mountain range, close to recharge areas where rare rainfall events help to chargerelatively freshwater aquifers. This water supply will help the cities of Abu Dhabi and Al Ain meet freshwater demands in the future.

SAUDI ARABIA

UNITED ARAB EMIRATES

QATAR

Doha

Abu Dhabi

Dubai

Sharjah

Qidfadesalination

plant

ShwaibASR site

OMAN

N

Al Ain

km

miles

75

75

0

0


Summer 2008 29

base maps. The data were compiled intogeographic information system (GIS) and HydroManager databases, then 3D hydrogeologicmodels were developed using Petrel and ECLIPSEgroundwater modeling and simulation software.

Water services experts from Schlumbergerworked in conjunction with EAD to identify andtest potential sites. They defined the storagezone, aquifer thickness and related hydraulicparameters, then tested potential candidateaquifers. The team ranked all potential ASR sitesto optimize the choice of locations. Exploratorydrilling strategies were developed for sites thatreceived the highest rankings, and wells werethen drilled. The wells were evaluated with acomprehensive suite of logs, including PlatformExpress integrated wireline logging tool, CMRand FMI tools and the ECS sonde.

As each well in the project area was developed,it was equipped with Diver data loggers tocontinuously monitor water levels, electricalconductivity and temperature, providing a reliablerecord of changes in response to source-waterinjection or hydraulic pumping tests. The datawere analyzed using AquiferTest Pro graphicaltest analysis system and KAPPA Engineering’sSaphir software to evaluate the hydraulic pro -perties of the aquifer. The results from this initialphase helped ASR project specialists choose an aquifer site that exhibited the requisite storage characteristics.

The test site is situated in the northeasternregion of the Emirate, southwest of Shwaib, on thewestern edge of the northern Oman Mountains.The site occupies an area of approximately 4 km2

[1.5 mi2], with unconsoli dated, quartz-rich sanddunes on the surface underlain by a 50-m [164-ft]thick unsaturated zone above about 25 m [82 ft] ofsaturated aquifer.

The aquifer consists of Quaternary unconsoli -dated eolian and fluvial sands, silty clays andcalcareous material deposited in paleochannelsthat were incised into Miocene mudstones andclaystones. The aquifer is bisected by a thrustfault that trends north to south and is dividedinto two units: an upper unsaturated, or vadose,zone and a lower saturated aquifer.11 It isunderlain by fine-grained Miocene sedimentaryrocks of the Fars formation, which has beensubdivided into Upper and Lower Fars units. TheUpper Fars unit consists primarily of claystonewith interbedded dolomitic marls, limestone andsiltstones. The Lower Fars unit is composedmainly of mudstone and evaporites.12

EAD oversaw the technical aspects of thisproject, which explored the hydrogeology andgeological structure of the site. The study calledfor injection of desalinated water into a shallowaquifer to determine water flow and mixingwithin the aquifer, followed by recovery andtesting of the water as it was pumped from thesubsurface (above). Microgravity measurementsand surface electromagnetic surveys were alsocarried out to further investigate the aquifer,determine the impact of thrust faulting and test

whether gravity techniques could be used todelineate aquifer boundaries.

10. Bradley CC, Ali MY, Shawky I, Lavannier A and Dawoud MA: “Microgravity Investigation of an Aquifer Storage and Recovery Site in Abu Dhabi,” First Break 25 (November 2007): 63–69.

11. The unsaturated zone, also referred to as a vadose zone,is typically the interval between the land surface and thetop of the water table. Pores within this zone containboth water and air. Pores in the saturated zone arealmost always completely filled by water.

12. Bradley et al, reference 10.

> Water discharged during a recovery test. Fresh water recovered from an ASR pilot test was pumpedmore than 2 km [1.2 mi] before being discharged into a dune depression. This water percolated about30 m [98 ft] vertically back into the aquifer. A nearby observation well monitored the process.


During this phase, two ASR wells were drilledin addition to a number of monitoring wellswithin the project boundaries. The test includedinjection-storage-recovery cycles, with a 30-daystorage period to assess potential hydrochemicalinteractions and to facilitate advanced waterchemistry analysis and the collection of moni -toring data from surrounding wells during thecycles (above). The project incorporated Petrelsoftware for geological modeling, ECLIPSE-H2Osoftware for dynamic groundwater simula tions,AquaChem manage ment, analysis and reportingsystem for water quality analysis, and PHREEQCfor water compatibility predictions.13 The HydroGeoAnalyst information management system andHydro Manager software provided complete

ground water and borehole data management,analysis and visualization.

While ASR provides a sound approach todeveloping and managing a long-term supply ofwater, it requires exceptional managementstrategies, backed by advanced aquifer-characterization capabilities and technicalexpertise. Until recently, problems with accuratecharacterization of the aquifer, water quality andrequired infrastructure have limited theapplication of ASR. This project tested andconfirmed the viability of ASR as a cost-effective,secure alternative to surface water storage. Thesite achieved a final system efficiency of 88%, and this project showed that 4 billion imperialgallons [18,184,367 m3] could be successfullystored at this location. Results of this study will

support development of the aquifer model, refineASR test planning and improve predictions forASR performance in this region. This ASRscheme also makes a vital contribution to thearea’s strategic water reserve. The government ofAbu Dhabi has adopted the ASR concept as asolution for strategic and seasonal storage offresh water.

30 Oilfield Review

13. PHREEQC is a computer program for speciation, batch-reaction, one-dimensional transport and inversegeochemical calculations, developed by the USGeological Survey. More information on this software is available at http://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc/ (accessed April 29, 2008).

14. Will RA, Yeh M, Eckhart L, Slade RC and Williamson MS:“Numerical Modeling of a Complex Regional AquiferRecharge and Recovery System (R3) in the UpperMojave River Basin, California,” (abstract)http://ngwa.confex.com/ngwa/expo07/techprogram/P4712.HTM (accessed May 19, 2008).

> Aquifer conditions during various stages of testing. A map view during a storage stage (top left ) shows the freshwater bubble of injected desalinatedwater (blue) and the buffer zone (transition from green to yellow to orange), surrounded by native water of the saline aquifer (red). The blue zonerepresents the freshwater bubble to be recovered. The cross section (bottom left ) shows ASR injection wells (red straws) and monitoring wells (greenstraws) penetrating an unconfined zone (gray) before reaching the aquifer. A freshwater bubble (blue) is surrounded by saline water (red). The 3Dvisualization (top right ) shows a freshwater mound developing during the injection phase. Later, a cone of depression is formed during the pumping phase(bottom right ). During this test, only a single well was used for injection and withdrawal.

Pres

sure

Low

High

N

ASR wellMonitoring well

Salin

ity

Fresh

Saline

Pres

sure

Low

High

N

Salin

ity

Fresh

Saline

N

ASR injector/recovery well

ASR injector well0.350.45

0.55

0.25

0.15

Monitoring well


Summer 2008 31

Mojave River Groundwater BasinThe Mojave River groundwater basin, situatedbeneath the Mojave Desert in southern California,involved an entirely different approach to aquiferrecharge. The proximity of this groundwaterbasin to the highly urbanized Los Angeles regionhas led to increased demand for its waters. TheMojave River, the primary source of surface waterfor the region, has only a small stretch ofperennial flow, with the remainder of its surfaceriverbed usually being dry—except for briefperiods of flow following intense storms.

The basin is managed in part by the MojaveWater Agency (MWA), a state water contractorthat, in cooperation with other California waterdistricts, is responsible for managing the region’swater resources to ensure sustainable watersupplies for present and future use. The MWA’sregional water management plan seeks tostabilize declining groundwater levels andaccommodate projected population growth. This

regional recharge and recovery (R3) programplays a major role in MWA’s long-term watermanagement strategy.14

When completed, the R3 project will let localwater providers use imported water tosupplement groundwater resources in this high-desert region of southern California. The R3

recharge project is being designed to utilizewater imported from northern parts of the statethrough a system of canals, aqueducts andpipelines operated by California’s State WaterProject, the nation’s largest state-built waterconveyance system. This imported water will bespread over the highly conductive, or permeable,sediment of the Mojave River floodplain, where itwill percolate down to the subsurface MojaveRiver aquifer. During times of increased demand,MWA will extract the stored water from a seriesof R3 replenishing wells, then route this waterthrough a secure conveyance system for deliveryto area water providers.

A key phase of the R3 program involvesconstruction of groundwater production anddelivery facilities, located down-gradient fromthe surface recharge zone. In light of the scopeand complexity of the project, the MWA deemedit necessary to develop high-quality predictivemodels for evaluating various recharge andrecovery system designs or testing operationalalternatives. The resulting models cover theupper Mojave River basin, the Mojave River andassociated surface recharge facilities.

A hydrogeologic conceptual model incor -porates detailed local structural andhydrostratigraphic features based on dataacquired by MWA and other agencies. This 3Dhydrogeologic model is divided into zonesrepresenting distinct strata interpreted fromgeological and geophysical data (below).

A transient model provides an accuraterepresentation of the Mojave River. While theriver is not the only key hydrologic feature in the

> Hydrogeologic model. An orthogonal section, based on a framework of satellite-measured digital elevation data, shows the upper Mojave Riverwatershed in California. The riverbed in the middle of the section cuts a gentle s-curve through alluvium shed from adjacent mountains. The cross-sectional view, built with the aid of well logs, is divided into major hydrostratigraphic horizons and shows the main aquifer level (yellow), bedrock andmountains (both purple). This model provides a framework for further transient simulations.

N

Geologic units

Surficial sediments

Unit 1

Unit 2

Unit 3

Unit 4Unit 5

Unit 6

Bedrock

X-axis

6,680,000 6,720,000 6,800,000 6,840,000 6,880,000 6,920,0006,760,000

0

2,000

4,000

6,000

8,000

Z-ax

is

1,960,000

2,000,000

2,040,000

2,080,000

Y-axis


model, it presents extreme challenges fortransient modeling, with its long periods of lowflow during dry periods and episodic flood flowsduring winter months. Mojave River inflow rateswere incorporated into the model developed bySchlumberger (above). This model helps theMWA study storm-event floods and groundwaterinteractions through the unsaturated zone.

The calibrated model allows civil engineers,hydrogeologists and municipal planners toinvestigate alternative R3 system-designscenarios to explore potential variations inrecharge location and timing, along withlocation, design and pumping schedules ofextraction wells. Evaluation constraints includeinfrastructure cost, impact on water quality,protection of endangered species and riparianareas, effects on existing municipal well fieldsand impacts on downstream users.

Once the infrastructure is in place, the R3

project will purify and store more than40,000 acre-feet, or 13 billion galUS [49 million m3],of imported water annually.

Groundwater GroundswellMore than 60% of the world’s population isconcentrated in coastal areas of the world—andthis trend continues to expand each day, withexpectations that it will eventually reach 75%. Asa result, water tables in some coastal communitiesare showing signs of over exploitation. The impactof this growth can sometimes be tasted in thewaters drawn from nearby aquifers. There,declining freshwater levels open the way forseawater intrusion and contamination ofunderground freshwater reserves.

Pumping from coastal groundwater reservoirscreates a low-pressure sink inland. If nopermeability barrier exists between thefreshwater aquifer and the coast, thenoverpumping of the aquifer will eventually drawsalt water inland, filling the aquifer andthreatening community water supplies. A similarproblem is seen in oil and gas wells. Excessivepumping on an oil or gas well can result in coningof subsurface waters, creating an adverse impacton production of desired fluids.

Coastal communities look to groundwatermanagers to combat this problem. On the westcoast of the United States, water authorities in LosAngeles County and Orange County, California,utilize some concepts similar to ASR to defendagainst encroaching seawater.15

Following years of heavy pumping to sustainregional agriculture, the water table had beendrawn below sea level, which allowed salt waterfrom the Pacific Ocean to encroach undergroundup to 5 miles [8 km] inland. To prevent furtherintrusion, the Orange County Water Districtdeveloped an artificial recharge program tocreate a subsurface hydraulic barrier. A blend of reclaimed water—5 million galUS/d[18,927 m3/d] reverse osmosis-treated water,9 million galUS/d [34,000 m3/d] carbonadsorption-treated water and 8.6 million galUS/d[32,555 m3/d] deep well water—is treated tomeet the state’s strict standards for drinkingwater. A series of 23 multipoint injection wells,located 4 miles [6.4 km] inland, delivers therecycled fresh water into coastal aquifers. Thedensity difference between fresh water and saltwater is used to good effect, as this “new” watercreates a freshwater buffer zone below ground.The buffer zone, also called a freshwater mound,pushes against the encroaching salt water toforce it away from municipal water wells, while monitor wells track the results of theinjection program.

With tools and technology in hand, it isimportant to address other aspects of thefreshwater challenge. Water shortages andpollution, in many cases, are societal problemsthat can be addressed, in part, by modifyingwater demand and usage. Although this remedyis certainly not as simple as it may sound, one approach toward this goal involves increasedawareness and education. Schlumberger istaking steps on this front as well.

Since 1998, the Schlumberger Excellence inEducational Development (SEED) program hasallowed Schlumberger employees, spouses andretirees to share their time and scientificexpertise with younger generations of learners.This global, nonprofit education programprovides access to knowledge and technological

32 Oilfield Review

15. For more on the Orange County Water District’s watertreatment and artificial recharge program, see the WaterFactory 21 discussion at http://www.ocwd.com/_html/wf21.htm (accessed April 29, 2008).

16. More information on these organizations can be foundon the Internet. To learn more about SEED initiatives:http://www.seed.slb.com/. The Web site for the MITMedia Lab can be accessed at http://learning.media.mit.edu/ (accessed April 29, 2008).

> Distribution of hydraulic conductivity. This map view of water-bearing units is divided into a riverfloodplain aquifer and regional aquifer. The river floodplain aquifer (orange) comprises hydraulicallyconductive sands and gravels deposited late in geological history when river flows were higher. Notsurprisingly, the floodplain traces the same gentle s-shaped curve seen in the previous figure. Thiscurve trends from orange to yellow with decreases in hydraulic conductivity. Hydraulic conductivityestimates from various sources were interpolated in 3D using a combination of deterministic andstochastic methods. Today, the river mostly flows underground. In the wider part of the orangesegment, the water table is quite close to the surface.

X-axis

6,680,000 6,720,000 6,800,000 6,840,000 6,880,000 6,920,0006,760,000

1,000

1

–0.001

Hyd

raul

ic c

ondu

ctiv

ity

(K),

ft/d

N


Summer 2008 33

resources for students ages 10 to 18, and for theirteachers in communities where Schlumbergerpersonnel live and work. The SEED programoffers a range of project-based activities, hands-on science education workshops and collab ora tiveinternational projects that typically focus onglobal themes such as water, climate change andalternative energy.

In observance of freshwater initiativesenacted by the United Nations, Schlumbergerlaunched the SEED Water Project to helpstudents and teachers develop the awareness,attitudes and skills essential to investigate waterquality and availability, maintain and improvethe quality of water sources in their localcommunities and to facilitate the sharing of theirdata and ideas with other students. In workshopshosted by SEED volunteers, students develop anunderstanding of water sources and their localwatershed, learn how to run accurate water-quality tests and explore the ways that water isregarded by different cultures (left).

SEED volunteers encourage students toinvestigate the health of local water sourcesusing low-cost water-quality testing kits. Thesekits, provided by the SEED project, give studentsthe tools for analyzing local water samples forpH, dissolved oxygen, biochemical oxygendemand, temperature, turbidity, nitrates,phosphates and coliform bacteria. SEED schoolsalso have the opportunity to post data on anInternet site, which provides a platform forcomparing data and reflecting on findings withina global context.

The SEED Water Project is facilitated bySEED staff and Schlumberger volunteers inpartnership with the Massachusetts Institute ofTechnology (MIT), and its Media Lab.16 Theefforts of SEED students and volunteers are part of a groundswell of environmentalstewardship that is gaining momentum aroundthe world. —MV

> Artwork from around the world. SEED participants from Ecuador, Kazakhstan,Mexico and India display their talents while celebrating water in their art.


34 Oilfield Review

Sand Injectites

Eric BracciniTotal E&P AngolaLuanda, Angola

Wytze de BoerMarathon Oil (United Kingdom) Ltd.Aberdeen, Scotland

Andrew HurstMads HuuseMario VigoritoUniversity of AberdeenAberdeen, Scotland

Gerhard TempletonMaersk Oil North Sea UK LimitedAberdeen, Scotland

For help in preparation of this article, thanks to Aimen Amer,Luanda, Angola; Robert S. Freeland, University of Tennessee,Knoxville, USA; Gretchen Gillis, Sugar Land, Texas, USA;Karen Sullivan Glaser and Matthew Varhaug, Houston; LarsHamberg and Cecilie Dybbroe Tang, DONG Energy, Hørsholm,Denmark; Patrice Imbert, Total E&P, Pau, France; Eric Jameson,Marathon Oil, Aberdeen; David McCormick, Josephine Ndinyahand Richard Plumb, Cambridge, Massachusetts, USA; David Mohrig, The University of Texas, Austin; Chris Murray,Pacific Northwest National Laboratory, Richland, Washington,USA; William Schweller, Chevron Energy Technology Company,San Ramon, California, USA; and Ian Tribe, Aberdeen.FMI (Fullbore Formation MicroImager), OBDT (Oil-BaseDipmeter Tool), OBMI (Oil-Base MicroImager), PeriScope, Q-Marine and UBI (Ultrasonic Borehole Imager) are marks of Schlumberger.

Sandstone dikes and sills, known as sand injectites, have long been considered

mere geological oddities. However, many operators are beginning to understand the

impact—both positive and negative—that injectites can have on E&P endeavors.

By using outcrop studies, core and log data and careful seismic illumination, companies

are now finding that some of these geological anomalies can be attractive exploration

targets and of huge significance when planning and optimizing hydrocarbon recovery.

> Common sand-injection features. Feature A is a depositional sand body that is also the parent bodyfor many of the injectites. B is a thick sill. C is a complex of thin sills and dikes. D is a set of sills thatlink with dikes in a stepping fashion. E is a large, irregular intrusive body that contains clasts of hostrock. F is a sill from Parent Body A that is crosscut by a dike from Parent Body J. G denotes sandextrusions and volcanoes. H represents gas seeps. I indicates conical sand injections. (Adapted fromHurst and Cartwright, reference 2.)

100 to 500 m

Modern seafloor

Ancient seafloor

A

B

C

D

E

F

G

G

H

H

20 to 100 m

J


Summer 2008 35

Sandstones and other clastic rocks are createdfrom wind- and waterborne sediments that settleunder the force of gravity. Younger sediments aredeposited on top of older ones. This observationis a cornerstone of geology.

Most sedimentary sequences retain thisarrangement unless faulting or folding placesolder above younger rocks. However, anothermechanism can disrupt the natural order;overpressured sediments can become remobil -ized and force their way upward, intruding intooverlying layers as the fluids carrying them seeklower pressures.

Structures formed by sediment injection arecalled injectites, or clastic intrusions. Becausethey resemble intrusive and extrusive igneousfeatures, much of the vocabulary for describinginjectites has come from igneous geology(previous page). Sills are emplaced parallel tobedding, whereas dikes cut through bedding. Thestrata containing the intrusion are called hostrock, and the layers that feed the intrusion arethe parent beds. In contrast, depositional bedsare those that have formed by settling oftransported sediment, not by injection. Sand-injection features exhibit size scales frommillimeters to kilometers, and have been seen incores, borehole image logs, seismic sections,outcrops, aerial photographs and satelliteimages. They have even been tentativelyidentified on photographs of Mars.1

E&P companies are learning that many oiland gas reservoirs in ordinary depositionalsandstones have connections to injectites. This isimportant for two main reasons. First, injectedsand can add volume to a reservoir, and add it inlevels that are structurally higher than the mainreservoir. Furthermore, injected sands typicallyhave high porosity and permeability, formingexcellent pay zones. Detecting the location andshape of injectites may help pinpoint additionalreserves and improve drilling success.

The second reason for interest in injectites ishydraulic communication. Injectites can improveconnectivity between reservoir layers, which mayallow reserves to be drained with fewer wells andless cost. However, increased connectivity canalso have a negative impact. The presence ofinjectites indicates a breach in the caprock thatnormally seals hydrocarbons within a reservoir. Ifa reservoir seal has been breached, the oil andgas might have escaped or be in communicationwith another reservoir. Injectites can also affectconnectivity with aquifers. These are importantconsiderations for optimizing field developmentand modeling reservoir behavior.

In this article, we review some of what hasbeen learned about injectites from outcropstudies and subsurface exploration. We alsodiscuss a few of the known hydrocarbonreservoirs currently believed to be associatedwith sand injectites and describe the impact ofinjectites on their development.

Early RecognitionClastic intrusions have been recognized inoutcrops, mines and subsurface data all aroundthe world (above).2 They have been described ingeological literature as far back as 1821.3 Duringhis voyage from 1832 to 1836 on the HMS Beagle,

Charles Darwin described a dike in easternArgentina as remarkable, slightly tortuous, andformed chiefly of rounded grains of quartz.4 Overthe next 170 years, geologists continued to reportthe presence of sandstone and other clastic dikesand speculate about their origin.5

In the early days of injectite observation, itwas believed that clastic intrusions could formonly if a large crack was open to the surface andfilled with sediments from above. Some clasticdikes, called Neptunian dikes, do form in thismanner, when extreme pressures from glaciersor other heavy depositional loads forcesediments down into underlying layers. It was

1. http://mars.jpl.nasa.gov/mgs/msss/camera/images/science_paper/f5/ (accessed July 11, 2008).

2. Hurst A and Cartwright J: “Relevance of Sand Injectitesto Hydrocarbon Exploration and Production,” in Hurst Aand Cartwright J (eds): Sand Injectites: Implications forHydrocarbon Exploration and Production, AAPGMemoir 87. Tulsa: AAPG (2007): 1–19.Ribeiro C and Terrinha P: “Formation, Deformation and Chertification of Systematic Clastic Dykes in a Differentially Lithified Carbonate Multilayer. SW Iberia,Algarve Basin, Lower Jurassic,” Sedimentary Geology 196,no. 1–4 (March 15, 2007): 201–215.Neuwerth R, Suter F, Guzman CA and Gorin GE: “Soft-Sediment Deformation in a Tectonically ActiveArea: The Plio-Pleistocene Zarzal Formation in the Cauca Valley (Western Colombia),” Sedimentary Geology 186, no. 1–2 (April 15, 2006): 67–88.Dharmayanti D, Tait A and Evans R: “Deep-Water Reservoir Facies of the Late Jurassic Angel Fan, DampierSub-Basin, Australia,” Search and Discovery Article 30044,posted November 4, 2006, http://www.searchanddiscovery.net/documents/2006/06127dharmayanti/index.htm(accessed May 21, 2008).

Chi G, Xue C, Lai J and Qing H: “Sand Injection and Liquefaction Structures in the Jinding Zn–Pb Deposit,Yunnan, China: Indicators of an Overpressured Fluid System and Implications for Mineralization,” EconomicGeology 102, no. 4 (June–July 2007): 739–743.Truswell JF: “Sandstone Sheets and Related Intrusionsfrom Coffee Bay, Transkei, South Africa,” Journal of Sedimentary Petrology 42, no. 3 (September 1972): 578–583.

3. Strangways WTHF: “Geological Sketch of the Environs ofPetersburg,” Transactions of the Geological Society ofLondon 5 (1821): 392–458. Cited in Newsom JF: “ClasticDikes,” Bulletin of the Geological Society of America 14(1903): 227–268.

4. Darwin CR: Geological Observations on South America.Being the Third Part of the Geology of the Voyage of theBeagle, Under the Command of Capt. Fitzroy, R.N. Duringthe Years 1832 to 1836. London: Smith Elder and Co. 1846.The Complete Work of Charles Darwin Online http://darwin-online.org.uk/content/frameset?viewtype=side&itemID=F273&pageseq=164 (accessed May 20, 2008).

5. Diller JS: “Sandstone Dikes,” Bulletin of the GeologicalSociety of America 1 (1889): 411–442.Newsom JF: “Clastic Dikes,” Bulletin of the GeologicalSociety of America 14 (1903): 227–268.

> Locations of clastic intrusions identified in outcrops, mines and subsurface data. (Adapted fromHurst and Cartwright, with additional data from Ribeiro and Terrinha; Neuwerth et al; Dharmayanti et al; Chi et al; and Truswell, reference 2.)

Injectite location


not until 1869 that a geologist proposed thatsediments could be intruded from below.6

An 1899 investigation of sandstone dikes innorthern California concluded that the intrusionsin question must have been formed by sand fillingfissures from below.7 The study noted that not allthe dikes reached the surface and includeddescriptions of fine-scale banding with parallelarrangements of mica plates and coarse and finesand along the sides of dikes. Some sandstoneintrusions also contained fragments of the hostrock. Geologists still use these characteristics,among others, to help distinguish clastic sills anddikes from depositional beds.

Another early observation on sandstone dikessuggested that oil could migrate through them toshallower reservoirs, or leak to the surface.8

Highly pressurized hydrocarbons have beenconsidered a possible source of the pressure andfluids responsible for injectite formation.9 Thisand other proposed causes of injectite formationare discussed later in this article.

Geologists and other E&P professionals arerevisiting surface exposures of injectites in thehopes of using them as analogs, or models, forinjectites encountered in the subsurface. Someof the larger outcrops, which appear to havespatial scales approaching those of majorsubsurface injectites, occur in the Panoche Hillsand near the Santa Cruz coast, both in California;the Magallanes basin, southern Chile; andTabarka, Tunisia. In this article, we describesome features of the injectites of the PanocheHills, what they suggest about injectite originsand how they can be used to better understandsubsurface injectites.

Outcrop ObservationsThe Panoche Hills are on the western edge ofCalifornia’s San Joaquin Valley. Sand-injectionfeatures were first recognized there in the early1900s, and have been studied by many groups.10

The vast network of sandstone sills and dikescrops out over an area greater than 350 km2

[135 mi2] and is observable in outcrop and aerialand satellite photographs (above left).

The sediments in the study area—the GreatValley sequence—were eroded from the SierraNevada mountains to the east during the LateJurassic and Cretaceous. These sediments, insome places 12 km [7.4 mi] or more thick, werelaid down in deep water as submarine fans andturbidites with interbedded siltstones andclaystones. Mud-rich sediments up to 1 km [0.6 mi] thick were deposited atop the sandyunits, creating a low-permeability seal.

36 Oilfield Review

> Sand injectites of the Panoche Hills, California. This network of light-colored sandstone dikes andsills emplaced in darker mudstone extends for another 700 m [2,300 ft] to the north. Apparent beddingis horizontal; sills are aligned horizontally and dikes crosscut bedding. The inset (bottom left ) showsan interpretation of dike and sill arrangement (black lines).

CaliforniaPanoche Hills

500 m

1,640 ft

North South

> Light-colored sandstone sills in darker mudstone. Sills with thicknessesup to 6 m were injected into mudstone and are linked by steps, essentiallydikes, that cut across bedding.


Summer 2008 37

In the Early Paleocene, a major injectionevent formed the giant Panoche injectitecomplex. This event emplaced fine- to medium-grained sands from Cretaceous submarine fansinto overlying mud-rich sediments. The injectitecomplex has a stratigraphic thickness of morethan 1,200 m [3,940 ft], and probably exceeded1,600 m [5,250 ft] before compaction.

Parts of the complex are dominated bystaggered sills—sills that rise from one level toanother in steps—at least 6 m [20 ft] thick(previous page, bottom). Some sills appear to befilled with mainly clean sand, while otherscontain rafts, or large inclusions, of clay-rich hostrock (right). Where sills have intruded, the hostrock is “jacked up,” showing apparentthickening. In portions of the complex, longdikes extend up to 1,200 m in length (belowright). In general, dike thickness decreases withdistance from the parent bed.

Some of the dikes reached the ocean floor,extruding sand onto the paleo-seabottom.Identification and dating of marine fossils alloweddetermination of an approximate timing of theinjection event. Isotopic analysis of calcitedeposited during fluid expulsion through theinjectites indicates that fluids seeped to the surfaceover a 2-million-year interval during the Danianstage, approximately 62 to 65 million years ago.11

6. Wurtz H: “On the Grahamite of West Virginia and the New Colorado Resinoid,” Proceedings of the AmericanAssociation of Science 18 (1869): 124–135. Cited in Newsom, reference 5.

7. Diller, reference 5.8. Newsom, reference 5.

Anderson R and Pack RW: “Geology and Oil Resources of the West Border of the San Joaquin Valley North of Coalinga, California,” US Geological Survey Bulletin 603 (1915).

9. Jenkins OP: “Sandstone Dikes as Conduits for OilMigration Through Shales,” AAPG Bulletin 14, no. 4 (April 1930): 411–421.

10. Anderson and Pack, reference 8.Jenkins, reference 9.Zimmerman J Jr: “Tumey Sandstone (Tertiary), Fresno County, California,” AAPG Bulletin 28, no. 7 (July 1944): 953–976.Payne MB: “Type Moreno Formation and OverlyingEocene Strata on the West Side of the San JoaquinValley, Fresno and Merced Counties,” California Divisionof Mines and Geology, Special Report 9 (1951).Smyers NB and Peterson GL: “Sandstone Dikes and Sills in the Moreno Shale, Panoche Hills, California,”GSA Bulletin 82, no. 11 (November 1971): 3201–3208.Friedmann J, Vrolijk P, Ying X, Despanhe A, Moir G andMohrig D: “Quantitative Analysis of Sandstone IntrusionNetworks, Panoche Hills, CA,” presented at the AAPG Annual Meeting, Houston, March 10–13, 2002.Vigorito M, Hurst A, Cartwright J and Scott A: “Regional-Scale Subsurface Sand Remobilization: Geometry andArchitecture,” Journal of the Geological Society 165, no. 3 (2008): 609–612.

11. Minisini D and Schwartz H: “An Early Paleocene ColdSeep System in the Panoche and Tumey Hills, CentralCalifornia, U.S.A.,” in Hurst A and Cartwright J (eds):Sand Injectites: Implications for HydrocarbonExploration and Production, AAPG Memoir 87. Tulsa:AAPG (2007): 185–197.

> A sandstone sill with large inclusions of host rock. The sill of light-colored sand contains large rafts, or inclusions, of darker mudstone hostrock that have been ripped up during intrusion. The inclusion nearest thegeologist has retained its horizontal orientation, but the inclusions to theleft have rotated. With this style of sill emplacement visible in outcrop, it iseasy to imagine that in the subsurface, large inclusions would have anunexpected and negative impact on wellbore stability and might be avoidedby acquiring the appropriate LWD measurements.

> Long dikes extending into the distance. These sandstone dikes (D)extend from the geologists in the foreground across several ravines andhills, approximately 1,200 m to the east (into the photograph). They aremore competent than the surrounding host rock, and so do not erode aseasily. The light-colored sill (S) in the foreground is the top of the sillcomplex seen in the photograph at the bottom of the previous page.

DD

D

S


The whole sequence was tilted toward theeast in the Paleocene during regional uplift thatalso formed the Coast Ranges to the west of theSan Joaquin Valley, at the same time as thedevelopment of the San Andreas transformmargin. Because of this tilt and subsequenterosion, the entire injection section, from parentrock to extrusion on the seafloor, can be seen in outcrop.

Another feature of the Panoche injectitecomplex is the presence of crosscutting dikes of different mineralogy (above). This indicates that several parent beds sourcedinjections independently during a single phase of sand injection.

The distribution and orientation of injectitesat the Panoche site provide some insight into thestate of stress at the time of sand intrusion. In

general, fractures open in planes perpendicularto the least principal stress. Therefore, wheresills dominate the injection style, the minimumstress direction was vertical. The alignment ofdikes over a great distance indicates that theywere emplaced when the least principal stresswas acting mainly in one horizontal direction.The presence of dikes in all directions, and ofdikes along with sills, indicates isotropic stressconditions.12 All these scenarios were active indifferent parts of the Panoche injectite complex.

Mechanics of Sand IntrusionThe mechanics of large-scale sand intrusion arenot well-known. One approach considersinjectites as natural examples of the inducedhydraulic fractures performed to stimulatereservoirs.13 With this approach, injection can bemodeled if the conditions of the injection eventare known or assumed. However, in most cases,neither the background conditions—such asfluid source, mode of sediment transport, depthand pore pressure of the parent sand, stressregime, and intrusion-emplacement depth andgeometry—nor the triggering mechanisms arewell understood. It also is not clear whetheroverpressured fluid initiates the fractures, whichare later filled with sand—analogous toproppant filling hydraulic fractures—or if thesand-laden fluid is the fracturing agent.

In spite of these limitations, there is somegeneral agreement about the three mainingredients required for generation of sandintrusions.14 The first is the occurrence ofunconsolidated sand encased in low-permea bilitymudstones. The size of the intrusion depends inpart on the amount of sand available. Small sand-rich channels can exhibit injection wings, but forlarge-scale intrusions, a greater volume of sandmust be present. Extremely large volumes ofinjected sand have been encountered—in somecases, 10 to 100 million m3 [350 million to 3.5 billion ft3].15 Also required are large volumesof fluid to transport the sand upward.

The second condition is overpressure causedby one or more mechanisms, such as disequi -librium compaction, lateral or deep pressuretransfer, fluid buoyancy and salt diapirism.Disequilibrium compaction arises when fluid-filled sand buried under low-permeabilitymudstone cannot expel pore fluids and compactnormally. Lateral pressure transfer, in the form oflarge-scale slumping, may impart overpressure toa buried sand body. Deep pressure transfer ariseswhen high overpressures from deep within asedimentary basin reach shallower levels.Migrating hydrocarbons, which are more buoyant

38 Oilfield Review

> Evidence of multiple parent beds. A golden-orange intrusion trendingfrom lower left to upper right is cut by a whiter intrusion that intersects itnearly perpendicularly at the feet of the geologist (arrow). Sand intrusionsof different colors indicate sources from multiple parent beds. Crosscuttingimplies multiple injection episodes.

> One set of well data, two models of sandstone distribution. Interpretation of well data depends onworking models. Log interpretation that originally predicted a distribution of thin, “ratty” sands abovethe main reservoir (top) may be modified if injectites are included in the interpretation (bottom).(Adapted from Hurst, reference 32.)

Ratty sands (depositional)

Blocky sands (depositional)“M

igration noise”

Differential compaction

“Migration noise”

Ratty sands(injected)

Blocky sands (injected)Sand wingSand

wing

Jacked-up host rock


Summer 2008 39

than water, could enter water-filled sand andincrease fluid pressure. Rising salt diapirs maycreate overpressure by pushing fluids ahead ofthem.16 Any of these mechanisms, alone or inconcert, could cause enough overpressure toinduce liquefaction, the condition in which thefluid—not the sand grains—bears theoverburden pressure, thereby reducing the shearstrength of the sand-fluid mixture to zero.

The third requirement is a triggering event,such as an earthquake, meteorite, comet orasteroid impact, volcanic eruption or landslide.These triggering events could induce a transientbuildup of overpressure, and also fluidization, orflow of the liquefied system.

Recognizing Subsurface Sand InjectitesSand injectites occur on size scales ranging frommillimeters to kilometers and typically juxtaposematerials with different physical properties.

These characteristics allow subsurface sand-injection features to be recognized in cores,wellbore image logs and seismic sections.

Injectites have been identified in cores fromoil fields in several areas, including the NorthSea, the Gulf of Guinea and offshore Australia.17

In cores, dikes are easier to recognize than sillsbecause of their clear discordance with host-rockbedding (above).18 For sills, in which theinjectite-host rock contacts are parallel,additional criteria, which also apply to dikes,may be used.19 These include homogenization, orlack of primary depositional structures, causedby fluidization. However, there may be internalstratification inconsistent with bedding andconsistent with remobilized sand, such as flowlaminae and alignment of grains or closerpacking of grains near injectite walls. Anotherindicator is the presence of host-rock clasts,which typically are angular. Fluid-escape

features, such as upward-merging structures,may be seen. In some cases, the intruding sandmay be affected by diagenesis, staining,cementation or residual oil in a manner that isdifferent from the host rock.

Because boreholes sample a relatively smallvolume of the subsurface, cores may undersampleinjectite volumes. If a borehole encounters aninjected sandstone, it is likely that there are moreinjectites nearby that have not been sampled.

Recognizing injectites in well logs often is notstraightforward. Injected sands generally do nothave a unique signature on resistivity or gammaray logs, and are frequently mistaken for thin, or“ratty,” sands (previous page, bottom). Therefore,one possible indicator of injectites is thepresence of thin sands above a massive sand body.Another sign is the presence of sand in unusualstratigraphic settings. Also, injectites tend to bethinner the farther they are from the parent sand.

12. Sand intrusion under isotropic stress conditions has alsobeen observed in Texas. For more: Diggs TN: “AnOutcrop Study of Clastic Injection Structures in theCarboniferous Tesnus Formation, Marathon Basin, Trans-Pecos Texas,” in Hurst A and Cartwright J (eds):Sand Injectites: Implications for HydrocarbonExploration and Production, AAPG Memoir 87. Tulsa: AAPG (2007): 209–219.

13. Jolly RJH and Lonergan L: “Mechanisms and Controlson the Formation of Sand Intrusions,” Journal of theGeological Society 159, no. 5 (2002): 605–617.

14. Huuse M, Cartwright J, Hurst A and Steinsland N:“Seismic Characterization of Large-Scale SandstoneIntrusions,” in Hurst A and Cartwright J (eds): SandInjectites: Implications for Hydrocarbon Exploration andProduction, AAPG Memoir 87. Tulsa: AAPG (2007): 21–35.

15. Hurst and Cartwright, reference 2.16. Marco S, Weinberger R and Agnon A: “Radial Clastic

Dykes Formed by a Salt Diapir in the Dead Sea Rift,Israel,” Terra Nova 14, no. 4 (2002): 288–294.

17. Braccini E and Penna E: “Sand Injections in AngolaDeep Offshore,” presented at the 5th Annual AngolaFormation Evaluation Forum, Luanda, Angola, October 26–27, 2005.Dharmayanti et al, reference 2.

18. Briedis NA, Bergslien D, Hjellbakk A, Hill RE and Moir GJ:“Recognition Criteria, Significance to Field Performance,and Reservoir Modeling of Sand Injections in the BalderField, North Sea,” in Hurst A and Cartwright J (eds):Sand Injectites: Implications for HydrocarbonExploration and Production, AAPG Memoir 87. Tulsa: AAPG (2007): 91–102.

19. Hurst A, Cartwright J and Duranti D: “FluidizationStructures Produced by Upward Injection of SandThrough a Sealing Lithology,” in Van Rensbergen P, Hillis RR, Maltman AJ and Morley CK (eds): SubsurfaceSediment Mobilization, Geological Society SpecialPublication 216. London: Geological Society (2003): 123–138.

> Cores with sand-injection features. The core on the left shows an injected sand with fluid-escape structures (courtesy of A.Hurst), which are subvertical tracks caused by fluid rising through unconsolidated sediments (inset ). The coin is approximately2 cm [0.8 in.] in diameter. The next core, from a Total E&P well offshore Angola, contains an oil-bearing sand dike (dark gray) in shale (light gray). The core on the right, also from Total E&P Angola, shows brecciated host rock (light gray) in injected sand(dark gray), along with a close-up view.

cm

0

10


Injected sand often is very well sorted andhas 30% or more porosity. Permeabilities arecommonly in the darcy range. However, in someinstances, injected sands have lower porosity,higher density and higher sonic velocities thantheir parent sands or other nearby depositionalsands.20 These criteria have been used todistinguish injectite from depositional facieseven when the injectite intrudes an unrelateddepositional sand.21

Borehole imaging tools, such as the FMIFullbore Formation MicroImager, the OBMI Oil-Base MicroImager and the UBI UltrasonicBorehole Imager, can detect sand injectites thatare discordant with host-rock bedding.Geologists at Total E&P Angola have used thesetools to image injectites in the Gulf of Guinea(left).22 Image logs provide an important linkbetween core-scale and larger-scale logmeasurements (below left).

At the seismic scale, sand-injection featuresare sometimes difficult to detect, because theyoften have a low acoustic impedance contrastwith the host-rock mudstones. Imaging of thesefeatures improves when contrast is high, as inthe case of high-porosity hydrocarbon-chargedsands juxtaposed with low-porosity, high-densityhost rock. For improved imaging in cases of lowacoustic impedance contrast, traditionallystacked compressional-wave processing may beaugmented with angle-stack processing,inversion and analysis of amplitude variationwith offset (AVO).23 The analysis of shear wavesobtained from ocean-bottom cable (OBC)acquisition can also reveal injection features notseen in compressional-wave data.24

The size of a feature that can be resolved byseismic methods depends on the densities andporosities of the layers through which theseismic waves travel, and on frequencybandwidth, spatial sampling, migration apertureand noise.25 In typical North Sea conditions,structures a few meters in size may be detected,and thicknesses in the range of 10 to 40 m [33 to131 ft] may be resolved, or quantified. Improve -ments in imaging achieved with application ofnew acquisition technology, such as the Q-Marinesingle-sensor system, are helping to resolve evensmaller features.26

A variety of injection styles has been observedin seismic data from the North Sea and off thewest coast of Africa. Injectite shapes can includewings, dipping structures that crosscut bedding,mounds, cuspate forms and cones (next page).27

40 Oilfield Review

> Sand injections in a UBI Ultrasonic Borehole Imager log. The UBI image(Track 3) shows sand injections from X,X02.5 to X,X03.0 m and from X,X04.0to X,X04.5 m (brackets) with high dips relative to the low dip of the host rock.The injectites sometimes correspond to slight decreases in gamma ray(green, Track 1). The high dip can also be seen in Track 1 (circle). Dips seenin the OBDT Oil-Base Dipmeter Tool image (Track 2) corroborate thoseinterpreted from the ultrasonic measurements (Adapted from Braccini andPenna, reference 17, courtesy of Total E&P Angola.)

OBDT Image UBI Image

0 0360 360

Dip, deg0 45 90

Gamma Ray0 200gAPI

X,X02.0

X,X03.0

X,X03.5

X,X04.0

X,X04.5

X,X02.5

Dep

th, m

> Correlating image logs and core in a Total E&P well in the Gulf of Guinea.Dips and images have been acquired with the OBMI Oil-Base MicroImagertool. Steeply dipping features corresponding to tadpoles circled in red(Track 1) indicate injections at high angles relative to the gently dippinghost rock. In the OBMI images (Tracks 2 and 3), dark colors correspond tolow resistivity (shales) and light colors correspond to high resistivity(sands). The colors of the sinusoids interpreted in the OBMI imagescorrespond to the colors of the dip tadpoles in Track 1. The corephotograph (right ) shows a high-angle contact between injected sand(light gray) and host rock (dark gray). (Adapted from Braccini and Penna,reference 17, courtesy of Total E&P Angola.)

OBMIStatic Image

0 360

OBMIDynamic Image

0 360Dip, deg

0 45 90

Gamma Ray0 200gAPI

20

30

40

50

60

70

80

Core

Vert

ical

sca

le, c

m

Dep

th, m

X,X50

X,X51

X,X53

X,X52

X,X54


Summer 2008 41

Oil Fields with InjectitesMany oil fields in the North Sea are linked toinjectites or contain sand-remobilization features.In some cases, thin intrusive sands above the mainreservoir have been identified through coresampling in exploratory wells but have beenignored or misunderstood. In other cases, seismicimaging helped identify large-scale injectionwings that were later found to contain hydro -carbons. Often these fields are centered on one orseveral large, deepwater sandstone unitsdeposited as turbidites or gravity flows. Thedistribution and size of injected sandsremobilizing from the parent sand bodies varywidely, as does the impact of sand injection onfield development.

The first North Sea oil discovery—Balderfield, in 1967—underwent more than 30 years ofexploration and appraisal before production.28

This geologically complex field comprises sevenoil accumulations in deepwater sands separatedand trapped by mudstones and volcanic tuff. In1969, centimeter-scale sand injections were firstrecognized in core from an early explorationwell, but were considered to be insignificant.29

Since then, more than 150 sand-injectionfeatures have been identified in core, log andseismic data of the Balder field, the largest witha thickness of 11 m [36 ft]. Postdepositional sandremobilization—sand withdrawal, diapirism andinjection—caused juxtaposition of sand on sand,creating widespread intrareservoir connectivity.Accumu lations in sand injectites may account formore than 25% of the oil in place in the Balderfield, and all the sands appear to be in pressurecommunication. In some areas of the field, it isimpossible to obtain reasonable history-matcheswithout adding a sand-injection component tointrareservoir connectivity.30 Incorporating sandinjectites in the full-field reservoir model helpedachieve sufficient matches with productionhistory and is expected to assist in maximizingtotal ultimate recovery from the field.

Perhaps the best-known example of a reservoiraffected by sand injection is the Alba field in theUK sector of the North Sea. The high-porosityturbidite channel sands of the Alba field werediscovered in 1984 while drilling for a deepertarget—the Brittania field. Production from Albafield began in 1994. As in the case of the Balderfield, sand-injection features were observed incores early on, but not considered important.

In 1998, a full-field 3D multicomponentseismic survey using OBCs revealed dippingreflections at the margins of the main channel.These structures had not been imaged by earlier,

20. Fretwell PN, Canning WG, Hegre J, Labourdette R andSweatman M: “A New Approach to 3D GeologicalModeling of Complex Sand Injectite Reservoirs: The Alba Field, United Kingdom Central North Sea,” in Hurst A and Cartwright J (eds): Sand Injectites:Implications for Hydrocarbon Exploration andProduction, AAPG Memoir 87. Tulsa: AAPG (2007): 119–127.

21. Duranti D, Hurst A, Bell C, Groves S and Hanson R:“Injected and Remobilised Eocene Sandstones from the Alba Field, UKCS: Core and Wireline LogCharacteristics,” Petroleum Geoscience 8, no. 2 (May 2002): 99–107.Hurst et al, reference 19.

22. Braccini and Penna, reference 17.23. McHugo S, Cooke A and Pickering S: “Description of a

Highly Complex Reservoir Using Single Sensor SeismicAcquisition,” paper SPE 83965, presented at SPEOffshore Europe, Aberdeen, September 2–5, 2003.

24. MacLeod MK, Hanson RA, Bell CR and McHugo S: “The Alba Field Ocean Bottom Cable Seismic Survey:Impact on Development,” The Leading Edge 18, no. 11(November 1999): 1306–1312.

25. Huuse et al, reference 14.Huuse M and Mickelson M: “Eocene SandstoneIntrusions in the Tampen Spur Area (Norwegian North

Sea Quad 34) Imaged by Seismic Data,” Marine andPetroleum Geology 21, no. 2 (February 2004): 141–155.

26. McHugo et al, reference 23.27. Molyneux S, Cartwright J and Lonergan L: “Conical

Sandstone Injection Structures Imaged by 3D Seismic in the Central North Sea, UK,” First Break 20, no. 6 (June 2002): 383–393.Davies RJ: “Kilometer-Scale Fluidization StructuresFormed During Early Burial of a Deepwater SlopeChannel on the Niger Delta,” Geology 31, no. 11(November 2003): 949–952.Hamberg L, Jepsen A-M, Ter Borch N, Dam G,Engkilde MK and Svendsen JB: “Mounded Structures of Injected Sandstones in Deep-Marine PaleoceneReservoirs, Cecilie Field, Denmark,” in Hurst A andCartwright J (eds): Sand Injectites: Implications forHydrocarbon Exploration and Production, AAPG Memoir 87. Tulsa: AAPG (2007): 69–79.

28. “History of the North Sea,” Norwegian PetroleumDirectorate, http://www.npd.no/English/Emner/Geografiske+omraader/Nordsjoen/NordsjoenHistorikk.htm (accessed July 8, 2008).

29. Briedis et al, reference 18.30. Briedis et al, reference 18.

> Seismic expressions of sand injection. Seismic imaging captures a range of injectite features.Mounded and cuspate forms appear in the top surfaces of injected sand (top left and top right ). A 3Dview (top left ) shows the top of a sand wing (blue) that rises steeply to the left. (Courtesy of MarathonOil UK.) At top right is a seismic section. (Adapted from Hamberg et al, reference 27, courtesy ofDONG E&P Exploration.) A conical intrusion in the shape of a “V” is seen above a thick depositionalsandstone (bottom left ). Vertical exaggeration is seven times. (Adapted from Huuse and Mickelson,reference 25.) An injectite with dipping wings that crosscut bedding is revealed in an inversionsection (center right ). (Courtesy of Marathon Oil UK.) A 3D view of a saucer-shaped injection from theNorth Sea (bottom right ) is color-coded from shallow (red) to deep (blue). (Courtesy of M. Huuse.)

~ 5 k

m

~1 km

Mounded Top Balder horizon overlying remobilizedsandstone mounds of the Siri Canyon


conventional towed-streamer surveys because ofthe poor acoustic impedance contrast betweensand and shale. Chevron geophysicists inter -preted the wing-like features as sand injections.31

Two wells were subsequently drilled into theinjectite wings; the first encountered 150 m[492 ft] of oil-bearing sand and produced20,000 bbl/d [3,178 m3/d].32 The second wellintersected 20 m [66 ft] of sand in the westernpart of the field.

The multicomponent seismic dataset hasbecome the primary tool for planning these wellsand predicting reservoir quality in the Alba field.33

Shear-wave impedance data from the invertedseismic volume—calibrated with core and logdata—have been converted to quantitative sand-quality values for the purposes of constructinggeological models for well placement andreservoir models for simulating production.

Several other fields of the North Sea exhibitinjectite and remobilization features, includingChestnut, Grane, Sleipner Øst, Volund, Gryphon,Leadon, Harding and Jotun. Two of these, theGryphon and Volund fields, are examined in thefollowing sections.

Sand Wings in the Gryphon FieldThe Gryphon field—Maersk Oil 86.5% equity,Sojitz 13.5% equity—was discovered in 1987 inBlock 9/18b of the UK North Sea (below). Thediscovery well penetrated 190 ft [58 m] of oil-bearing sand in the Balder formation at a depth of5,700 ft [1,738 m].34 The main Gryphon reservoir,on production since 1993, comprises basin-floorturbidite sandstones of Eocene age. However,since 2004, operator Maersk Oil has also beenproducing oil from wells drilled into seismic-scalesand-injection wings extending from the mainreservoir. These features are developed at themargins of the Gryphon field where the Baldersand depositional system terminates.

Like most fields now known to be associatedwith sand-injection features, the Gryphon fieldwas initially thought to contain only depositionalsand formations. As more evidence wascollected, the interpretation evolved. Earlyexploration and appraisal drilling in the 1980sand 1990s revealed a complex distribution ofreservoir sands. For example, one well near thecenter of the field encountered more than 300 ft[91 m] of turbidite sand, while a well less than1,640 ft [500 m] north encountered almost nosand. Reservoir quality was excellent, with

porosity averaging 36% and permeabilityaveraging 7 D.35 Adding to reservoir complexity,small-scale injected sands, up to a fewcentimeters in thick ness, were seen above themain reservoir, but were not considered tocontribute signifi cantly to the reservoir volume.

Early seismic investigations—2D surveys in1985, 1987 and 1988, a 3D survey in 1990 and anOBC survey in 1999—detected the broadlymounded Gryphon structure, but limitations inthe seismic data quality made detaileddelineation difficult. Later, after Maersk Oilgained experience with injectites in the nearbyLeadon field, together with improvements inseismic processing and simultaneous inversionapplied by Maersk Oil to the long-offset seismicdata acquired in 2002, large-scale sand-injectionwings were identified on the edges of the field(next page, top right).36

The first injectite development target was asection of the major sand wing on the easternedge of the field. This target involved numerouschallenges including geosteering through adipping sheet-like sand body and managingwellbore stability in high-porosity, unconsol-idated sands. The uncertainty in the seismicallyderived position of the sand wing, possiblycaused by limitations in migrating its steeperdip, together with depth-conversion discrepancy,has an impact on the lateral position of a dippingsand wing. This proved to be a subsurface teamchallenge in geosteering along strike in thedipping sand wings.37

These difficulties were overcome by a teamapplying a combination of tools, such asprototype geosteering technology—Maersk Oilwas the first company in the UK sector of theNorth Sea to use PeriScope bed boundarymapper technology and to drill a dedicatedproduction well in an injection-wing target—anddetailed prewell scenario planning, includingwellbore-stability studies and extensive inter -disciplinary collaboration. An office-basedgeosteering team operated around the clock andcommunicated with the rig crew to integrateLWD data with the understanding of the injectitereservoir, enabling Maersk Oil geologists to makereal-time decisions to geosteer back into thesand wing after exiting the injectite to maximizethe amount of pay along the wellbore.

Wellbore-stability studies using data fromearlier Gryphon wells that penetrated the Balderformation—about one quarter of them hadexperienced mud losses—indicated that well -bore stability could not be achieved. Eitherbreakouts or mud losses could be managed, but

42 Oilfield Review

> Gryphon field, UK North Sea. Since 2004, the Gryphon field has been producing oil from horizontalwells geosteered into injected sands.

9/18b-34Z

9/18b-A26

9/18b-A25

9/18b-A27

9/18b-A28

9/18b-A28Z

9/18b-A23

9/18b-A24

9/18b-A24Z

2004

2005

2007

Year Wells Drilled

10 miles

0 1km

Gryphon Field

9/18b-A23

9/18b-A24Z

9/18b-34Z

9/18b-A26

9/18b-A25

9/18b-A27

9/18b-A28Z

VolundGryphon

Alvheim

1000 miles

0 100km

NORWAY

UK

N O R T H S E A


Summer 2008 43

not both. Drillers adopted a zero-loss strategy andworked toward managing borehole failure byoperating within a strict equivalent circulatingdensity (ECD) window.38 With the additional riskof exiting sand and encountering unstable shales,successful drilling required following provenpractices, such as adherence to swab and surgelimits during drilling, connections and tripping.

In 2004, the first Gryphon development wellto target an injection wing was drilledhorizontally for 1,420 ft [433 m] along the strikeof the wing, penetrating sand along 53% of itslength. LWD measurements indicated that theborehole trajectory was on the low side of therugose sand wing rather than in its center,explaining the multiple sand entries and exits.However, while the wellbore was within theinjection wing, the net-to-gross ratio was 100%.

The second well was designed to enteranother segment of the major sand wing on thewestern edge of the Gryphon field and tocontinue horizontally along the strike of thewing. The well entered the wing as expected, butexited through the base into the surroundinghost shale. Returning to the 12¼-in. boreholewith the PeriScope 15 bed boundary mapperconfirmed the position of the wellbore relative tothe sand wing.39 The PeriScope 15 tool usesazimuthal induction measurements to sense

resistivity contrasts up to 15 ft [4.5 m] from theborehole for real-time geosteering.

An openhole sidetrack was drilled 80 ft[24 m] to the west, landing in the sand wing asdesired (below). The sidetrack largely remainedwithin the injection wing for 1,440 ft [439 m],penetrating sand along 80% of a 1,800-ft [549-m]horizontal well section.40 Again, while thewellbore was within the injection wing, the net-to-gross ratio was 100%.

With the two successful wells drilled in 2004,production from Gryphon field more than

doubled, reaching 27,000 bbl/d [4,290 m3/d] by theend of that year. These results encouraged Maerskto drill additional wells, incorporating newknowledge about seismic positioning uncertainty.

In the 2005 drilling campaign, a lateral shiftwas applied to well paths to account for theuncertainty in seismic positioning. Also, newwells were planned to land a few hundred feetabove the wing and make a gradual, obliqueapproach; once the wing was found, casing wasset and the first operation in the 8½-in. sectionwas to veer along the strike of the wing. With

31. MacLeod et al, reference 24.Lonergan L and Cartwright JA: "Polygonal Faults andTheir Influence on Deep-Water Sandstone ReservoirGeometries, Alba Field, United Kingdom Central NorthSea," AAPG Bulletin 83, no. 3 (March 1999): 410–432.

32. Hurst A: “Sand Intrusions Reveal Increased Reserves,”GEO ExPro (October 2005): 12–20.

33. Fretwell et al, reference 20.34. Purvis K, Kao J, Flanagan K, Henderson J and Duranti D:

“Complex Reservoir Geometries in a Deep Water ClasticSequence, Gryphon Field, UKCS: Injection Structures,Geological Modelling and Reservoir Simulation,” Marine and Petroleum Geology 19, no. 2 (2002): 161–179.

35. Templeton G, McInally A, Melvin A and Batchelor T:“Comparison of Leadon and Gryphon Fields SandInjectites: Occurrence and Performance,” presented atthe 68th EAGE Conference and Exhibition, Vienna,Austria, June 12–15, 2006.

36. For more on simultaneous inversion: Barclay F, Bruun A,Rasmussen KB, Camara Alfaro J, Cooke A, Cooke D,Salter D, Godfrey R, Lowden D, McHugo S, Özdemir H,Pickering S, Gonzalez Pineda F, Herwanger J,Volterrani S, Murineddu A, Rasmussen A and Roberts R:“Seismic Inversion: Reading Between the Lines,” Oilfield Review 20, no. 1 (Spring 2008): 42–63.

37. Hart N, Ageneau G, Mattson P and Fisher A:“Development of the Gryphon Field Injection Wing—Technical Challenges and Risks,” paper SPE 108655,presented at SPE Offshore Europe, Aberdeen,September 4–7, 2007.

38. Hart et al, reference 37.39. Chou L, Li Q, Darquin A, Denichou J-M, Griffiths R,

Hart N, McInally A, Templeton G, Omeragic D, Tribe I,Watson K and Wiig M: “Steering Toward EnhancedProduction,” Oilfield Review 17, no. 3 (Autumn 2005): 54–63.

40. Hart et al, reference 37.

> Seismic interpretation of injection wings in the Gryphon field. Thereservoir is outlined in yellow. The injection wings are visible as dippingfeatures on the right side. The gas-, oil- and waterprone intervals of thereservoir are shaded light red, green and blue, respectively. Gamma raylogs (black) are displayed along well trajectories (orange).

Top Balder

A22

Massive sand

A2ZA19Z A24Z

Injection wings

Top sand

> Landing a horizontal well in a sand wing. The first attempt to hit this injected sand overshot andcame out the other side. This successful well landed as planned, and stayed within the wing for 1,440 ft. LWD logs (bottom right ) helped drillers stay in the sand-injection wing. The PeriScope 15 image(bottom) shows that the well stayed mostly in the high-resistivity sand (yellow) but did encounter afew patches of low-resistivity claystone (brown). The sand and claystone zones are also evident inthe gamma ray (green) and resistivity (red) curves.

Wellbore

Resistivity

Gamma Ray and Caliper

PeriScope 15 Interpretation

X,300 X,600 X,900 Y,200 Y,500 Y,800

Wate

Gas

r

Top ofwing

Base ofwing

Oil

Tens of feetthick

Wellboretrajectory


these improvements, three additional wells weredrilled in 2005, and two more in 2007.

These wells encountered a range of reservoirqualities in different portions of the injectite. For example, the first well in 2005 was drilledinto a large, continuous sand that was easilynavigated, while the second well was challegingto geosteer and penetrated an overall loweramount of sand along hole (above).

By the middle of 2008, the seven wells thattargeted sand-injection features had produced14 million barrels [2.22 million m3] of oil and aredelivering approximately 80% of Gryphon field’sdaily production (below).41

Exploiting the injection-wing play has provedvery successful for Maersk Oil. The Gryphon fieldhas been rejuvenated, with further sand-wingtargets identified for drilling, and more

importantly, the subsurface injection-wingexpertise is being applied to other areas operatedby Maersk Oil in the North Sea.

Discovering Oil in a Sand-Injection ComplexA remarkable example of successful injectiteexploration is the giant sand-injection complexof the Volund field (formerly called Hamsun),discovered by Marathon and partner LundinNorway AS in 2004 in Norwegian Block 24/9. Thisdiscovery, believed to be the world’s firstdeliberate drilling into an injection feature notconnected to a producing field, containsestimated reserves of 40 to 50 million barrels[6.4 to 7.9 million m3] of oil equivalent.42

Before the Marathon partnership acquiredthe license in 2003, six exploration wells had beendrilled in Block 24/9 in the Norwegian sector ofthe North Sea (next page, top left). Wells 24/9-5and 24/9-6, drilled by Fina Exploration Norway in1993 and 1994, found minor oil columns in latePaleocene and Early Eocene mounded sands, butthese were not of economic interest.43

As operators elsewhere in the North Seabegan recognizing and developing sand-injection features associated with producing oil fields, interpreters revised their assessmentsof the seismic data covering the marginal oilfinds in Block 24/9. The basin-shaped structurewas identified as a clastic-intrusion complex with steeply dipping sands that could be highly connected.44

In an independent evaluation, Marathonreprocessed the 1996-vintage 3D seismic data tosee if offset-dependent amplitude variations

44 Oilfield Review

> PeriScope 15 LWD images from Gryphon wells drilled in 2005. The first well (top) encountered 1,733 ft[528 m] of continuous sand that was easily navigated (light color). The second well (bottom) penetrated673 ft [205 m] of a thinner sand of lower quality with several low-resistivity claystone zones (dark color).The green lines are the planned well paths, and the red curves are the actual well paths.

500 1,0000

True

ver

tica

l dep

th, f

t

5,700

5,750

5,800

Horizontal length, ft

500 1,0000 1,500

True

ver

tica

l dep

th, f

t5,700

5,750

Horizontal length, ft

41. Hart et al, reference 37. 42. http://www.marathon.com/Global_Operations/

Exploration_and_Production/Norway/ (accessed July 2, 2008).

43. De Boer W, Rawlinson PB and Hurst A: “SuccessfulExploration of a Sand Injectite Complex: HamsunProspect, Norway Block 24/9,” in Hurst A andCartwright J (eds): Sand Injectites: Implications forHydrocarbon Exploration and Production, AAPG Memoir 87. Tulsa: AAPG (2007): 65–68.

44. Lawrence DA, Sancar B and Molyneux S: “Large-ScaleClastic Intrusion in the Tertiary of Block 24/9, NorwegianNorth Sea: Origin, Timing and Implications for ReservoirContinuity,” presented at the AAPG InternationalConference and Exhibition, Birmingham, England,September 12–15, 1999.Huuse M, Duranti D, Guargena C, Prat P, Holm K,Steinsland N, Cronin BT, Hurst A and Cartwright J: “Sand Intrusions: Detection and Significance forExploration and Production,” First Break 21 (September 2003): 33–42.

45. De Boer et al, reference 43.46. In a Class III AVO case, the sand has a lower acoustic

impedance contrast than the encasing shale and has alarge negative reflection coefficient at normal incidence.For more: Rutherford SR and Williams RH: “Amplitude-Versus-Offset Variations in Gas Sands,” Geophysics 54,no. 6 (June 1989): 680–688.

47. Quakenbush M, Shang B and Tuttle C: “PoissonImpedance,” The Leading Edge 25, no. 2 (February 2006):128–138.

> Sand-injection impact on production. Oil from the seven horizontal wells inGryphon field’s sand wings has revitalized production. Injection-wing wells nowaccount for about 13% of cumulative production from the field.

Oil

rate

, bbl

/d

0

10,000

20,000

30,000

40,000

50,000

60,000

Main field wells

Injection-wing wells 13%

87%

Cumulative fieldproduction

October 1993 October 2007


Summer 2008 45

might reveal information about fluid content inthe injected sands (above right).45 The increasedamplitudes on the far-offset stacked sectioncompared with those of the near offsets—known as a Class III AVO signature—are typicalof a hydrocarbon-filled, high-porosity sand in this area.46

However, it was important to increaseconfidence in the nature of the hydrocarbon andto confirm the likelihood of an oil-filled structure.Because developing a gas accumulation was not

economically viable at the time, furthergeophysical assessment of fluid type wasnecessary to limit risk before committing to drillthe prospect. Marathon geophysicists performedprestack 3D inversion for shear-wave and

compressional-wave impedances, which werecombined to yield sections of so-called Poissonimpedance (below).47 Inversion enhanced theinterpretability of the seismic data, clearlydelineating the 3D shape and internal structure

> Volund field, south of Alvheim field in Block 24/9 of the Norwegian NorthSea. Wells 24/9-5 and 24/9-6 (labeled 5 and 6, respectively), drilled in 1993and 1994, found insignificant amounts of oil. The labeled seismic line showsthe position of the seismic sections in the next two figures on this page.

VolundGryphon

Alvheim

1000 miles

0 100km

NORWAY

UK

N O R T H S E A

56

7A

7 7B

7C

10 miles

0 1km

Volund Field

Seismic line

> Near-offset and far-offset seismic sections through the Volund sand-injection complex. Because of the particular contrast in acoustic propertiesbetween the injected sand and the host rock, the dipping wings of theinjectite are better imaged in the far-offset section (bottom) than in thenear-offset section (top). The bottom of the injection complex is interpretedin green. The top of the Balder shale is interpreted in yellow and is seen to“jack up,” or thicken, discontinuously in the center, where the injectedsands are thickest. The increased amplitudes on the far-offset stackedsection compared with those of the near offsets are typical of ahydrocarbon-filled, high-porosity sand in this area. Gamma ray logs(yellow) are displayed along well trajectories (black).

Volund Near-Offset Amplitudes

24/9-7A24/9-7

Balder shale

Volund Far-Offset Amplitudes

24/9-7A24/9-7

Balder shale

> Inversion of Volund field seismic data. Inversion results for Poisson impedance depend on the background model used. An initial model with flat layersparallel to bedding (left ) produces an image of injection wings with poor continuity. An initial model that includes the “bathtub” shape of the injectite (right )yields sand wings with better continuity in their dipping sections and makes it easier to distinguish the injected sand from the host rock.

Poisson Impedance Inversion, Flat Model

24/9-7A24/9-7

Poisson Impedance Inversion, “Bathtub” Model

24/9-724/9-7A


of the injection feature (above). Calibration ofPoisson impedance values with well log dataallowed differentiation of sand-rich units fromsurrounding mudstones. The seismic inversionimages also revealed a flat spot, whichcorresponded to the oil/water contact (OWC)observed in Well 24/9-6, increasing confidencethat the sand injectite was filled with oil.

Analysis of outcrop analog data from centralCalifornia, including observations from thePanoche Hills injectites, gave Marathon inter -preters confidence in the size and geometry ofinjected sand bodies in the North Sea. Inoutcrop, clean injected sands of all sizes wereseen to be well-connected with each other andalso with poorer quality breccia-laden sands.Applying this concept to the subsurface prospectmade it likely that sand intervals of high net-to-gross values would be in hydrauliccommun i cation with zones of lower net-to-grossvalues, and that any hydrocarbon migrating intothe system would fill the structure and be inpressure communication.48

The first exploration well, Well 24/9-7,targeted the southern wing of the injectitecomplex and encountered an injected sand layerat a depth of 1,848 m [6,063 ft], within 2 m[6.6 ft] of the predicted reservoir top. Asexpected from the seismic interpretation, theborehole penetrated two major intrusive sanddikes. The upper one, unexpectedly containinggas, had a true vertical thickness of 32 m [105 ft],and the lower one, an oil-bearing zone, had a truevertical thickness of 12 m [39 ft]. Core recoveredfrom this well contained host-rock mudstone,injected sand and breccia with host-rock clasts(above right).

Following the discovery of oil in the first well,sidetrack Well 24/9-7A was drilled downdip tofollow the gas-filled dike into the oil column. Thetwo sand intrusions were again intersected, theupper one in the oil leg and the lower one now inthe water leg. Another sidetrack, 7B, probedupdip into the fringe of the injectite complex,

46 Oilfield Review

48. De Boer et al, reference 43.

> Core section from Well 24/9-7. The two mainreservoir facies are captured in this section: fine-grained sandstone (light gray) and breccia withdark, angular mudstone clasts. The mudstone atthe bottom of the section is a large clast.

0.6

m

Brecciainjectite facies

Injected sand

Mudstonehost rock

> Volund field injection-complex top, interpreted from seismic inversion data. This view is from westto east. Surface colors indicate depth; red is shallow and blue is deep. The line of the previouslyshown seismic section is white. Wells are black lines. Vertical exaggeration is approximately fourtimes. Wing height is 250 to 300 m [820 to 984 ft] and maximum wing dip is 25 to 30 degrees.

1 km0.62 mi

North


Summer 2008 47

and found several thin, gas-filled, centimeter-scale sands (above). This proved important forseismic calibration purposes. A third sidetrack,7C, drilled into the eastern portion of theinjectite complex, intersected a 49-m [160-ft]massive sand and penetrated the OWC in thelocation that was predicted by extrapolation ofdata recorded in Well 24/9-6. All the sand in theinterval corresponding to the Paleocene age isthought to be injected from the underlyingHermod parent sand.

The clean sands in the Volund reservoir haveporosities averaging 30% and permeabilities ofseveral darcies. Reservoir quality decreases with

height; thick massive sands in the deeper portionof the structure give way to thinner, predom -inantly low net-to-gross breccias toward theedges of the injection complex. Estimatingreserves and optimizing development of the fieldrequire predicting the 3D distribution ofreservoir facies within the intrusion structure.The seismic inversion results combined with logdata and rock physics analysis can be used asguides for generating various scenarios of faciesdistribution (below).

Development plans for the Volund fieldinclude three horizontal producing wells and onewater-injection well. Marathon geoscientists

have made continued use of outcrop analoginformation in planning these wells, foreseeingthe use of real-time LWD measurements togeosteer and anticipating borehole stabilityproblems when drilling through the injectedsand dikes.

Development drilling and production fromthe Volund field are expected to begin during2009. The field will be developed as a subseatieback to Alvheim field 10 km [6 mi] away.

> Correlation of logs from exploration and appraisal wells. Wells (top) are displayed at equaldistances, not at true horizontal locations. For each well, gamma ray is displayed with lithology, andresistivity is displayed with fluid content to the right and left of each well, respectively. Red is gas andgreen is oil. Sand content decreases upward within the wing. The Balder tuff and Balder shale arejacked up, showing apparent thickening to accommodate the injected sand. The inset (right ) indicatesthe location of the cross section relative to the injectite complex.

Hordaland group

Heimdal

Lista

Sele

Balder tuff

Balder shale

Injected sands

Hermod

24/9-6 24/9-7A 24/9-7 24/9-7B

7

7B

7A

6

> Distribution of reservoir facies within the Volund intrusion structure. A combination of seismic inversion results with rock physics analysis leads to anunderstanding of the distribution of connected breccia (left ) and connected sand (center). Their sum produces the total connected reservoir volume (right ).


Effects of InjectitesKnowing from the start that a reservoir isaffected by clastic intrusions has a significantimpact on field development. The excellentconnectivity typical of injectites allows reservoirsto be developed with fewer wells. One operatingcompany acknowledged that a reservoir offshorewest Africa could have been developed with halfthe number of wells if the extent of connectivitycaused by injectites had been known in advance.Operators in several areas are benefitting fromthe experiences of North Sea pioneers byincorporating injectite-related connectivity intheir development-drilling plans.

Evidence of hydrocarbon migration throughfluid-injection structures can be seen inoutcrop. For example, oil-filled sand intrusionscrop out on the beach north of Santa Cruz,California (left). These large sills are sourcedfrom sand deeper in the section, which traveledto the surface through dikes and then filled with hydrocarbon.

The additional connectivity caused by sandinjectites must be considered not only foroptimal production, but also when reservoirs areput to other uses. For example, the Sleipner Østfield in the Norwegian North Sea is a potentialsite for future storage of gas from nearby wellsafter it has exhausted producible reserves.49 Thenine reservoir zones of high-quality sand areseparated by mud-prone zones that can becorrelated throughout the field. However,pressure data show that the sands are inpressure communication. Sandstone dikes andsills have been seen in cores from several wells,and may be responsible for the enhanced verticalpermeability seen in the pressure data. OperatorStatoilHydro believes that assessment of theeffects of injectites on reservoir behavior will becritical in the optimization of recovery during thelate-stage development.

Another application in which clasticinjectites may have an effect is waste storage.For instance, the Hanford site in the state ofWashington has been used by the US Departmentof Energy for storage of contaminants in surfacetanks.50 Developing plans for eventual closure ofthe tank farms requires accurate models for fluidtransport, including the effect of clastic dikes onfluid flow.51 Thousands of clastic dikes arepresent at the surface and below the Hanford site(left). Understanding how this network of fine-grained dikes behaves during periods of low andhigh fluid flow is key to optimizing the design ofremedial systems that may be needed.

48 Oilfield Review

> Clastic dike in Hanford, Washington. The fine-grained dike (brown) wasemplaced in fine- to medium-grained host rock (gray). The dike is 0.7 m [2.3 ft] wide on the lowest face. (Photograph from Murray et al, reference 51.)

> Oil-filled outcrop north of Santa Cruz, California. The sandstone (dark)between two mudstones (light) originated from parent sand layers deeper inthe section and rose through nearby dikes (not shown). Oil, now biodegradedand heavy, causes the darkening of the sandstone. Large inclusions of light-colored mudstone can be seen encased in the sandstone.


Summer 2008 49

Other near-surface studies have helpedidentify some of the mechanisms that triggersediment injection.52 Investigations of sandinjectites emplaced during historically docu -mented earthquakes show a clear connectionbetween seismic triggering and widespreadoccurrence of clastic dikes, sand “blows” andother fluid-escape features.53

For example, in 1811 and 1812, the area nearNew Madrid, Missouri, USA, experienced threelarge earthquakes—each with magnitude 8 orhigher. During these seismic episodes, sandysediments in the shallow subsurface wereliquefied and vented to the surface over an areaexceeding 3,600 mi2 [9,320 km2].54 Extruded sandand associated dikes can be seen in excavations

(left). These seismically induced features can beused to infer the location and strength ofpaleoearthquakes. This information is importantfor designing buildings and other structures inareas that experienced seismic activity predatingwritten records.55

Sand Intrusions ElsewhereThe marked impact of sand injection in the NorthSea leads many explorationists to wonder ifinjectites are influential in other basins.Although literature on injectites is sparse forother areas, the reported occurrences in the Gulfof Guinea and offshore northwest Australiaindicate that sand-injection features are playinga role beyond the North Sea.

Other major oil-producing regions, such asoffshore Brazil and the Gulf of Mexico, report fewor no occurrences of injectites, and this leadsgeologists to wonder why. In the Gulf of Mexico,only a few cores have revealed finger-sized sandintrusions. Some experts believe that injectitesexist, but haven’t been identified. Others arguethat sedimentary conditions there, such as claymineralogy and sand grain-size distribution, arenot conducive to injectite formation.

Although clastic injectites have not beenconsidered significant in the Gulf of Mexico, theregion exhibits other fluid-escape features, in theform of shallow water flows that expel fluid at theseafloor. Fluid escape through shallow waterflows might be precursory to sediment extrusionon the seafloor. Fluid and sediment expulsiononto the seafloor can undermine surfaceinstallations, and the overpressured units fromwhich they spring make drilling and completingwells difficult.56

To better understand the conditions thatcause shallow water flows, an expedition of theInternational Ocean Drilling Program investi -gated overpressured sediments in the area of theUrsa field in the Gulf of Mexico.57 High-resolutionseismic data showed no obvious sand-injectionstructures, but may have detected incipient dikeformation. Pressure measurements in over -pressured formations indicated that fluidpressures are high, but insufficient to fracturethe overburden, and so not yet ready to form sand injectites.

Although the Gulf of Mexico and other areasare not yet known to be significantly affected bysand injections, as more geoscientists considerthe possible presence of injectites wheninterpreting data, injectites will be identifiedmore often. Recognizing injection features incore, logs and seismic data requires anawareness of these features, a perceptive mindand a trained eye. —LS

> Excavated sand dike. This sand dike, associated with the 1811 and 1812 New Madrid earthquakes,fills a fissure that is 200 ft [61 m] long. (Photograph courtesy of Carl Wirwa, Tennessee WildlifeResources Agency, Alamo, Tennessee, reference 53.)

49. Satur N and Hurst A: “Sand-Injection Structures inDeep-Water Sandstones from the Ty Formation(Paleocene), Sleipner Øst Field, Norwegian North Sea,” in Hurst A and Cartwright J (eds): Sand Injectites:Implications for Hydrocarbon Exploration and Production,AAPG Memoir 87. Tulsa: AAPG (2007): 113–117.

50. “Hanford State of the Site 2007 Meetings,” http://www.hanford.gov/?page=651&parent=0 (accessed July 22, 2008).

51. Murray CJ, Ward AL and Wilson JL: “Influence of ClasticDikes on Vertical Migration of Contaminants in theVadose Zone at Hanford,” Pacific Northwest NationalLaboratory Report PNNL-14224 prepared for theDepartment of Energy, March 2003.

52. Obermeier SF: “Seismic Liquefaction Features: Examplesfrom Paleoseismic Investigations in the ContinentalUnited States,” USGS Open File Report 98-488. http://pubs.usgs.gov/of/1998/of98-488/ (accessed July 8, 2008).

53. Obermeier SF: “The New Madrid Earthquakes: An Engineering-Geologic Interpretation of RelictLiquefaction Features,” USGS Professional Paper 1336-B, 1989.

54. http://web.utk.edu/~freeland/projects/sb.htm (accessedJune 12, 2008).

55. Obermeier SF: “Use of Liquefaction-Induced Features for Seismic Analysis—An Overview of How SeismicLiquefaction Features Can Be Distinguished from OtherFeatures and How Their Regional Distribution andProperties of Source Sediment Can Be Used to Infer theLocation and Strength of Holocene Paleo-Earthquakes,”Engineering Geology 44, no. 1 (1996): 1–76.

56. Myers G, Winkler C, Dugan B, Moore C, Sawyer D,Flemings P and Iturrino G: “Ursa Basin Explorers ShineNew Light on Shallow Water Flow,” Offshore Engineer(September 2007): 88–93.

57. Moore JC, Iturrino GJ, Flemings PB, Hull I and Gay A:“Fluid Migration and State of Stress Above the Blue Unit,Ursa Basin: Relationship to the Geometry of Injectites,”paper OTC 18812, presented at the Offshore TechnologyConference, Houston, April 30–May 3, 2007.


50 Oilfield Review

Moving Natural Gas Across OceansS. Andrew McIntoshBP Trinidad and TobagoPort of Spain, Trinidad

Peter G. NobleJim RockwellConocoPhillipsHouston, Texas, USA

Carl D. RamlakhanAtlantic LNG Company of Trinidad and TobagoPoint Fortin, Trinidad

For help in preparation of this article, thanks to Michelle Foss,University of Texas, Austin; and Patricia Ganase, Atlantic LNG,Point Fortin, Trinidad.Coselle is a mark of Sea NG Corporation. Invar is a mark ofImphy Alloys. Moss is a mark of Moss Maritime. OptimizedCascade is a mark of ConocoPhillips.

Significant natural gas reserves are located in remote areas that lack a local market

and where pipeline transport may not be economical. Increasingly, that gas is

converted to the liquid phase and sent to import terminals around the globe. Liquefied

natural gas is at the forefront of growth in low-emission, clean-energy sources.


Summer 2008 51

Natural gas has come a long way since 390 BCE,when the Chinese used it in the manufacture ofsalt. In the 2,400 years since then, its scope hasexpanded substantially—from simple saltproduction to transport across oceans in theform of liquefied natural gas (LNG). In the past100 years, the use of natural gas has grown froma local fuel to a regional resource and is nowpoised to become a global commodity.

Although the early Chinese, Romans andGreeks made limited use of gas as an energysource, wider use did not occur until about 1800with the introduction of town gas derived fromcoal for lighting.1 Demand for natural gas grewduring the early part of the 20th century, but itsuse remained primarily local until soon afterWorld War II. Engineering technology developedat that time was used to construct safe and reliable long-distance pipelines for naturalgas transport.

As natural gas evolved from local to regionaluse, applications expanded from home fuel topetrochemical feedstock to electric power

generation. Gas consumption for power gen era -tion surged during the last 25 years with theintroduction of efficient gas turbines and therecognition of the inherent environmentalbenefits associated with natural gas. Today,electric power generation accounts for morethan half of the growth in gas demand. The USEnergy Information Administration has esti -mated that world gas consumption will grow by70% between 2002 and 2025.2

Although natural gas consumption isexperiencing rapid growth, gas finds have notalways been looked upon favorably by theirdiscoverers. During much of the 20th century,natural gas markets were constrained by lowprices and oversupply. Gas that could not be soldwas flared or reinjected in gasflood projects todisplace oil or maintain reservoir pressure.Those attitudes have changed because ofincreased emphasis on pollution control.

Natural gas is the cleanest burning fossil fuel.Potential emission levels of sulfur, nitrogen andparticulates from natural gas are orders of

magnitude lower than are those from oil or coal.Although refiners and power plants can clean upmuch of the emissions from oil and coal, theyspend significant energy and capital to do so. Inaddition to low pollutant emissions, combustionproducts from natural gas contribute signifi -cantly less greenhouse-gas emissions. Naturalgas combustion emissions of carbon dioxide[CO2] are 40% less than those of oil and 80% lessthan coal, based on energy content.3

In recognition of its favorable emissioncharacteristics, natural gas has been called the“fuel of the future,” and its use is now equivalent

1. The Chinese transported gas in bamboo pipes fromshallow wells to gas-fired brine evaporators for makingsalt. For more information: Kidnay AJ and Parrish WR:Fundamentals of Natural Gas Processing. Boca Raton,Florida, USA: Taylor & Francis Group, 2006.Town gas is a flammable vapor made by heating coalwith steam. It is a mixture of carbon monoxide, hydrogen,methane and volatile hydrocarbons. For more information:http://www.123exp-technology.com/t/03884354486/(accessed June 8, 2008).

2. Tusiani MD and Shearer G: LNG. Tulsa: PennWellPublishing Company, 2007.

3. “Natural Gas and the Environment,” http//www.naturalgas.org/environment/Naturalgas.asp (accessed May 3, 2008).


to coal as an energy source.4 This status must betempered by the disparity between where naturalgas is found and where it is consumed (below).5

On the resource side, 60 to 70% of worldwidenatural gas reserves are in six countries with overhalf of that in Iran and Russia.6 Among consumingareas, the United States and the European Unionaccount for nearly 50% of gas use.7 In addition tothe mismatch between reserve and consumptionlocations, about 60% of reserves are consideredstranded.8 Stranded natural gas reserves have no local market, and transport by pipeline is not economical.

Locations where stranded gas cannot bemoved by pipeline present a limited number ofalternatives. One option is gas-to-liquids (GTL)technology in which natural gas is converted tohigh-quality liquid hydrocarbons by the Fischer-Tropsch reaction.9 The basic chemistry for thisprocess was developed in Germany in the earlypart of the 20th century and has been the focusof significant research to improve the catalystsand reactors used. Although several GTL sitesare in operation, the technology is complex,plants are expensive, and the stranded gasvolumes used as feedstock must be large enoughto justify the capital expenditure.

Another option is marine transport in theform of compressed natural gas (CNG).10 CNG isa solution for connecting small gas reserves withsmall markets over intermediate distances.Although GTL technology and CNG will answerneeds in some markets, currently the most prac -tical solution for moving large volumes of naturalgas over long transoceanic distances is LNG.

The reason for liquefying natural gas issimple. At atmospheric pressure, as natural gasis cooled to form LNG, its volume decreases by afactor of about 600. This decrease in volumemakes it economically attractive to liquefy andtransport gas from large stranded fields todistant consumers. What distinguishes LNG frommost other oilfield liquids is that it is cold—near−160°C [−256°F] at its boiling point and atatmospheric pressure.11 This liquid form ofnatural gas is pumped to specially designedmarine carriers for transport to terminals thatare often thousands of miles away. The chain ofliquefaction plants and import terminals indifferent parts of the globe linked by seagoingtransport is referred to as the LNG value chain.12

The costs related to each part of the valuechain are high, and in the past, LNG projectswere associated only with long-term contracts.13

Higher energy prices are changing the LNGmarket. The emergence of spot trades andmovement of cargos to distant rather than closerimport terminals indicate that LNG has becomea global commodity.14

The focus of this article is LNG—how it isliquefied, transported and stored until it isregasified for the consumer. Examples demon -strate the technology involved at each step in theLNG chain, including steps taken to ensure LNGsafety. Also discussed is the impact of largerliquefaction plants and vessels on the way theindustry views terminal location.

Liquefaction—The First StepLiquefaction has a long history. Although theBritish chemist and physicist Michael Faraday isbest known for his work in electricity, during theearly part of the 19th century he also liquefiednatural gas.15 This work was followed by that ofKarl von Linde and David Boyle who built thefirst practical refrigerators in the 1870s.16 In thelast part of that century, Linde developed aprocess to make commercial quantities of liquidoxygen and nitrogen.

52 Oilfield Review

> Natural gas reserves and consumption. The largest natural gas reserves are primarily located in the Russian Federation and the Middle East. TheRussian Federation has proven reserves of 44.7 trillion m3 [1,577 Tcf], while reserves in Iran, Qatar, Saudi Arabia, the UAE and the USA total 72.6 trillion m3

[2,563 Tcf]. Smaller but still significant reserves are found in Iraq, Nigeria, Venezuela, Algeria and Indonesia. These proven reserves equal 21.1 trillion m3

[745 Tcf], while the remaining global reserves of 39 trillion m3 [1,377 Tcf] are spread across 42 countries. The largest natural gas consumer is the UnitedStates at 653 billion m3/yr [23.1 Tcf/yr] followed by the Russian Federation at 439 billion m3/yr [15.5 Tcf/yr] and Iran at 112 billion m3/yr [3.9 Tcf/yr].

Iraq Iran

SaudiArabia

UAE

Canada

USA

Venezuela

UK Germany

Italy

Algeria

Nigeria

Russia

JapanChina

Indonesia

Qatar

Proven natural gas reserves,

5 trillion m3

Annual natural gas consumption,

70 billion m3


Summer 2008 53

The work of these early pioneers in gas lique -faction led to an interest in liquefying natural gasas a method for storing it compactly. The firstLNG facility was constructed in 1912 in WestVirginia, USA, and the first commercial plant wasbuilt in Cleveland, Ohio, USA, in 1941.17 EarlyLNG plants such as the Cleveland facility wereused for peak shaving. Utilities use peak shavingto supplement the supply of natural gas duringperiods of high demand.18 During periods of lowdemand, the LNG reserves are replenished.

Liquefaction plants built to process gas fromstranded natural gas reserves are termed

baseload plants and now constitute the bulk ofLNG capacity. One of the first baseload plantswas built in 1969 by ConocoPhillips at Kenai,Alaska, USA, to process natural gas from fields atCook Inlet. LNG from this plant is still beingexported to the Japanese power market.

Since construction of the Kenai plant, LNGliquefaction capacity has grown steadily, but notalways smoothly. Plans generated during the oilcrises of the 1970s were abandoned during theglutted markets of the 1980s. The currentbackdrop of high prices, tight supply and desirefor clean fuels has spurred growth in LNG

liquefaction capacity. In the past decade,capacity has doubled from 86 million tonUK/yr[94.8 million tonUS/yr] to about 183 milliontonUK/yr [201.7 million tonUS/yr], andliquefaction plants can be found in all parts ofthe world (above).19

Despite their cost and complexity, lique -faction plants are simply large refrigerators, andrefrigeration is at the heart of the LNGliquefaction process. Enough heat must beremoved from the natural gas to take it fromambient conditions to about −160ºC. In a closed-loop refrigeration process, a refrigerant

4. Fesharaki F, Wu K and Banaszak S: “Natural Gas: TheFuel of the Future in Asia,” http://www.eastwestcenter.org/fileadmin/stored/pdfs/api044.pdf (accessed June 9, 2008).Makogon YF and Holditch SA: “Gas Hydrates as aResource and a Mechanism for Transmission,” paperSPE 77334, presented at the SPE Annual TechnicalConference and Exhibition, San Antonio, Texas,September 29−October 2, 2002.

5. “Statistical Review of World Energy 2008”,http://www.bp.com/productlanding.do?categoryId=6929&contentId=7044622 (accessed July 11, 2008).

6. The six countries are Russia, Iran, Qatar, Saudi Arabia,the UAE and the USA. For more information: Kidnay andParrish, reference 1.

7. “International Energy Outlook 2007,” http:www.eia.doe.gov/oiaf/ieo/nat_gas.html (accessed May 21, 2008).

8. “A Dynamic Global Gas Market,” Oilfield Review 15, no. 3(Autumn 2003): 4−7.

9. “Turning Natural Gas to Liquid,” Oilfield Review 15, no. 3(Autumn 2003): 32−37.

10. Stenning S and Mackey T: “CNG Opens New Markets,”Fundamentals of the Global LNG Industry. London:Petroleum Economist (2007): 67−68.

11. The normal boiling point of pure methane is −162°C[−259°F]. Pipeline gas destined for LNG production mustbe treated to remove impurities that might freeze duringliquefaction. Residual amounts of hydrocarbons andother gases remaining after pretreatment leave LNGwith a boiling point slightly above that of pure methane.For more information: http://encyclopedia.airliquide.com/Encyclopedia.asp?GasID=41 (accessed June 11, 2006).

12. Tusiani and Shearer, reference 2.13. The total cost for all components in the LNG chain (gas

reserve development, liquefaction, vessels and importterminal) is about US $4 to 6 billion.

14. Davis A and Gold R: “Surge in Natural-Gas Price Stokedby New Global Trade,” The Wall Street Journal CCLI,no. 91, April 18, 2008.

15. “Brief History of LNG,” http://www.beg.utexas.edu/energyecon/lng/LNG_Introduction_06.php (accessedMay 16, 2008).

16. “Karl von Linde Biography (1842−1934),” http://www.madehow.com/31inventorbios/31/Karl-von-Linde.html(accessed May 15, 2008).

17. Foss MM: “Introduction to LNG,” http://www.beg.utexas.edu/energyecon/lng/Documents/CEE_Introduction_To_LNG-Final.pdf (accessed May 4, 2008).

18. Peak-shaving LNG plants combine three elements—liquefaction, storage and regasification. In 2004, the USAhad 59 peak-shaving plants. For more information:Kidnay and Parrish, reference 1.

19. Chabrelie MF: “LNG, The Way Ahead,” Fundamentals ofthe Global LNG Industry. London: Petroleum Economist(2007): 10–14.

> LNG liquefaction plants. Baseload liquefaction plants are found on every continent except Antarctica and cluster in regions with large stranded gasreserves—North Africa, the Middle East and Australasia. There are 20 baseload LNG liquefaction plants in operation with four of those undergoingexpansion. Six baseload plants are in the construction phase. Plants at Snǿhvit in Norway and Sakhalin in Russia illustrate a trend toward operation inharsh, arctic environments.

In operation

Under construction

21502schD7R1.qxp:21502schD7R1 10/6/08 9:44 PM Page 53

com pound is used to cool a natural gas or someother fluid.20 The refrigerant must pass throughseveral stages before returning to the startingpoint and beginning the process again. Thesestages are called refrigeration cycles and aremost often illustrated on Mollier charts.21 Sincerefrigeration is invariably accompanied by alarge energy input, cycles for industrial refri -geration try to approach the ideal Carnot cycle asclosely as equipment and operating procedureswill allow.22 Many oilfield operations use a simplevapor refrigeration cycle with Joule-Thomsonexpansion (below).23

While the Joule-Thomson cycle is mostappropriate for simple refrigeration tasks, it hassometimes been used to produce LNG commer -cially. For example, an isolated utility nearVancouver, British Columbia, Canada, built a smallLNG plant 64 km [40 mi] away to supply fuel forpower generation.24 Trucks moved LNG betweenthe two facilities. At the plant, the inlet gas wascompressed to 20.7 MPa [3,000 psi] beforeundergoing two stages of irreversible Joule-

Thomson expansion—first to 2.1 MPa [300 psi],then to a final gauge pressure of 0.07 MPa [10 psi]to produce LNG. For this small facility, the designsimplicity of the Joule-Thomson expansion out -weighed the thermo dynamic inefficiency of anirreversible process. In contrast, current commer -cial LNG plants seek to minimize the temperaturedifference between the natural gas being cooledand the refrigerant. This is accom plished bytailoring the refrigerant and employing more thanone stage. These considerations overwhelm anydesign simplicity in the Joule-Thomson expansion.

LNG liquefaction plants are large, complexprocessing facilities. These plants include threeseparate areas—feed-gas cleanup, liquefactionand storage plus vessel loading. Because of theextremely low temperatures present in LNGproduction, typical pipeline gas must undergoextensive cleanup before liquefaction. Impurityremoval in an LNG plant is designed to addressthree potential problems.25 First, contaminantssuch as water and carbon dioxide are aggres -sively removed to prevent freezing during

liquefaction, which would plug lines and otherequipment. Secondly, nitrogen can raise thepotential for stratification in LNG tanks, and itsconcentration is typically reduced to less than1 mol%. Finally, mercury is removed to a levelbelow 0.01 µg/m3. Higher levels of mercurycorrode the aluminum in liquefaction heatexchangers, ultimately causing failure.

Following cleanup, the treated natural gasenters the liquefaction section of the plant.Design of liquefaction processing is driven by adesire to approach the ideal Carnot-engineefficiency. That efficiency of 100% occurs whenthe process is entirely reversible and the coolingcurve of the material being refrigerated and theheating curve of the refrigerant correspond toone another exactly.26 Although this level ofefficiency can be achieved only for an ideal case,current LNG plants have made significantprogress in approaching it. Keys to the efficiencyof these plants are found in three areas—refrigerants, compressors and heat exchangers.

Today’s LNG plants offer two generalrefrigerant alternatives—mixed refrigerant andpure-component cascade.27 For example, the C3-MR process—developed by Air Products &Chemicals—uses propane and a multi componentrefrigerant to liquefy treated natural gas to LNGin two refrigerant cycles. An alternative tech -nology—the ConocoPhillips Optimized Cascadeprocess—includes three pure-component refrig -erant cycles to progressively cool and liquefy thenatural gas (next page). Each approach hasadvantages and disadvantages, and the ultimatechoice depends heavily on customer and siterequirements. In 2006, about 80% of worldwideLNG capacity employed the mixed refrigerantprocess, with the remaining 20% using pure-component cascade liquefaction.

If liquefaction is the heart of the LNGprocess, then compression and the associatedgas turbine drivers provide the muscle. Bothcentrifugal and axial compressors are used tocompress the refrigerant.28 Mixed-refrigerantcompressors must handle high capacities at lowtemperatures and often use axial compressors.On the other hand, the ethylene refrigerant—used in the ConocoPhillips cascade process—might require a centrifugal compressor.Compressor type and design depend on theparticular refrigerant and service. Centrifugalcompressor efficiency for LNG plants built in the1970s was nearly 70%. Current efficiencies forcentrifugal compressors in the same service are80% or greater.29 In current plants, thesecompressors are driven by gas turbines, and LNG

54 Oilfield Review

> Vapor refrigeration cycle. A common use of Joule-Thomson expansion in the oil industry is thesingle-stage propane refrigeration cycle. This cycle has four components—compressor, condenser,expansion valve and evaporator. Operation of this system consists of four steps that can be visualizedon a propane Mollier pressure-enthalpy diagram. The cycle begins at Point A with propane as asaturated vapor at atmospheric pressure and −40°C [−40°F]. The propane vapor is compressed to anabsolute pressure of 1.62 MPa [235 psi] and 93°C [200°F] at Point B. Condensation of the vapor fromPoint B at constant pressure to a 49°C [120°F] saturated liquid at Point C takes place by heatexchange with an external coolant—typically air. The propane liquid at Point C is expanded through aJoule-Thomson valve arriving at Point D as liquid-vapor mix at −40°C and atmospheric pressure. Thefinal step takes the propane liquid-vapor mix at Point D through an evaporator where it loses its latentheat by cooling an external stream. The propane ends where it began, as a saturated vapor at Point A.

Pres

sure

, psi

Enthalpy, Btu/lbm

–100 –50 0 50 100 150 200 250 300 350

100

200

80

60

40

20

10

8

6

4

2

1

400

600

800

1,000

AD

BC

Compressor

Evaporator

Condenser

Joule-Thomsonexpansionvalve

Sat

urat

ed li

quid

Sat

urat

ed v

apor

Critical point


Summer 2008 55

20. Refrigerants vary depending on the nature of the systembeing cooled. Propane is widely used as a refrigerant inoilfield applications, while various hydrofluorocarboncompounds are used as living-space refrigerants.

21. “Mollier Charts,” http://www.chemicalogic.com/mollier/default.htm (accessed June 26, 2008).

22. The Carnot cycle for a heat engine consists of fourstages (two isothermal and two adiabatic). Since all ofthe processes in a Carnot cycle are reversible, there isno change in entropy, making it the most efficient cycle.Thermodynamic cycles—such as the Carnot cycle—thatcan only be approached but never actually realized areoften termed “ideal” cycles. For more information:http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/carnot.html (accessed June 10, 2008).

23. In a Joule-Thomson expansion, a refrigerant isirreversibly expanded through an orifice or throttlingvalve. For more information: Kidnay and Parrish,reference 1.Smith JM and Van Ness HC: Introduction to ChemicalEngineering Thermodynamics. New York City: McGraw-Hill Company, 1975.

24. Blakely R: “Remote Areas of Canada Can Now BeServed by Trucked LNG,” Oil & Gas Journal 66, no. 1(January 1968): 60−62.

25. Kidnay and Parrish, reference 1.Tusiani and Shearer, reference 2.

> LNG liquefaction process alternatives. LNG may be liquefied using amulticomponent refrigerant in a single cycle or using several purecomponents in a cascade arrangement. The multicomponent or mixedrefrigerant approach is typified by the Air Products & Chemicals C3-MRprocess (top left ). Dry, treated natural gas is precooled to about −30°C[−22°F] by propane to remove liquid propane and other natural gas liquids.The precooled gas is sent to the main cryogenic heat exchanger where it is condensed and then flashed to produce LNG at −160°C. After heatexchange to produce LNG, the mixed refrigerant—typically nitrogen,methane, ethane, propane, butane and pentane—is sent to compression to complete the cycle. This process can achieve an average approach—

the temperature difference between refrigerant and material beingcooled—of about 8.3°C [15°F] (top right ). The ConocoPhillips OptimizedCascade process uses three separate pure-component refrigerant cyclesto produce LNG (bottom left ). Similar to the mixed refrigerant process, thecascade process first uses a propane refrigerant loop to remove naturalgas liquids from the treated natural gas. That material is then subjected totwo more refrigerant cycles—ethylene and methane—that produce theresulting LNG. Each refrigerant loop consists of independent compressors,expansion valves, condensers and evaporators. This process can achievean average approach of about 6.7°C [12°F] (bottom right ).

100

0

–100

–200

–3000 20 40 60 80 100

Enthalpy change, %

Gas being liquified

Refrigerant

100

0

–100

–200

–3000 20 40 60 80 100

Enthalpy change, %

Gas being liquified

Refrigerant

Heavy

Compressor

Mixed refrigerant

Expander

Propane

Air-fin heatexchanger

Separator

Compressor Turbine Compressor Compressor Turbine


Light

LNG

Heavy-component removal

Natural gas liquids

Trea

ted

natu

ral g

as

Plant fuel

LNG

Air Products C3-MR Mixed-Refrigerant Process

EthylenePropane


Air-fin heatexchanger Plant fuel


LNG

Trea

ted

natu

ral g

as

Methane

Natural gas liquids

LNG

Vapors fromvessel loading

ConocoPhillips Optimized Cascade Process

Compressors Turbines TurbinesCompressors Compressors Turbines

Heavy-component removal

26. These curves are often called duty curves and plottemperature versus enthalpy (heat content). For more information: Ransbarger W: “A Fresh Look atProcess Efficiency,” LNG Industry (Spring 2007),http://lnglicensing.conocophillips.com/publications/index.htm (accessed July 26, 2008).

27. Although there are several variants, the majority ofmixed refrigerant LNG plants use the Air Products &Chemicals technology. Similarly, ConocoPhillipstechnology dominates use of pure-component cascade.

28. “Liquefied Natural Gas—Enhanced Solutions for LNGPlants,” http://www.geoilandgas.com/businesses/ge_oilandgas/en/downloads/liquified_natural_gas.pdf(accessed June 13, 2008).

29. Ransbarger, reference 26.


production capacity is directly related to thepower that these turbines can deliver. A plantbuilt in 2000 might use gas turbine units thatenabled LNG production of 3.3 million tonUK/yr[3.7 million tonUS/yr], while current large gasturbines are associated with capacities of almost7.8 million tonUK/yr [8.6 million tonUS/yr].30 Likecompressors, the turbines used in LNG plantshave seen an increase in cycle efficiency—from28 to 40% during the last 30 to 40 years.

The final key to efficient natural gas lique -faction is effective heat transfer. Specializedheat transfer equipment ensures that thetemperature difference between refrigerant and

the natural gas being cooled is minimized. Mostof the heat transfer equipment used in currentLNG liquefaction plants has grown out of otherefforts in the cryogenics industry. Generally,three types of specialized heat exchangers areused in LNG refrigeration circuits—plate-fin,spiral-wound and core-in-kettle.31

Aluminum plate-fin exchangers—used in thecascade process—consist of alternating layers offins and plates enclosed in a rectangular vesselshell.32 Compared with comparable carbon orstainless-steel equipment, plate-fin exchangersare 20% of the size and 10% of the weight. Incontrast, a spiral-wound exchanger is the

primary heat exchange device used to produceLNG in the mixed refrigerant process.33 Small-diameter tubing is wound around a central coreand this tube assembly is inserted in a pressurevessel shell (below left). Both the cascade andmixed refrigerant processes may use the thirdtype of exchanger—core-in-kettle. This exchangeris usually used in propane heat exchange andconsists of a plate and fin-type block placed withina large horizontal cylindrical shell.34

It is typical for LNG plants to have multipleprocessing lines, or trains, for liquefying naturalgas. This allows for the planned expansion of thefacility. The output from single or multiple trainsis sent to insulated storage tanks nearby wherethe LNG remains until a vessel arrives for loadingand shipment to distant terminals.35

Caribbean Gas Moves GloballyThe LNG plant at Point Fortin, Trinidad, uses themodern liquefaction technology described above.Large natural gas reserves lie offshore theCaribbean island of Trinidad, and over the past50 years this gas has been used for electricitygeneration, methanol and ammonia production—and now LNG export. The Atlantic LNG Companyof Trinidad and Tobago was formed in 1995 todevelop an LNG plant at Point Fortin.36

The Atlantic plant at Point Fortin wasconstructed on 838,000 m2 [207 acres] ofreclaimed land to process gas from offshore gasfields southeast and north of Trinidad (next page,top left). The plant began operation in 1999 witha capacity of 3.0 million tonUK/yr [3.3 milliontonUS/yr] of LNG and 950 m3/d [6,000 bbl/d] ofnatural gas liquids. Following the success of theinitial operation, expansion projects werelaunched in 2000 and 2002 to increase the plantcapacity from one to four trains capable ofproducing a total of 14.8 million tonUK/yr[16.3 million tonUS/yr] of LNG and up to3,820 m3/d [24,000 bbl/d] of natural gas liquids.37

In 2007, Atlantic LNG Company had the seventhlargest LNG production capacity in the world andwas the largest supplier of LNG to the USA (nextpage, top right). In the past, most of thecompany’s product moved to import terminals inthe USA, but that is changing. Because of highergas prices in Europe, significant productshipments are now sent to terminals in Spain. Asa consequence, Atlantic LNG Company is playinga key role in Atlantic basin gas pricing.38

All Atlantic LNG trains at Point Fortin usethe ConocoPhillips Optimized Cascade process.When the first train was built in 1999, it was the

56 Oilfield Review

> Spiral-wound and plate-fin heat exchangers. Mixed refrigerant liquefaction uses a spiral or coil-wound design (left ) as the main cryogenic heat exchanger. In this device, small-diameter tubes arewrapped around a center core—called the mandrel—in alternating directions. In the design shown,the fluid in the tubes enters at the bottom and moves upflow before leaving at the top. The stream onthe vessel side passes downflow over the tubes yielding counterflow heat exchange between thefluids. Each tube terminates in tube sheets that are part of the cylindrical shell. For LNG service, an all-aluminum design is typical, and heat transfer surface/volume ratios of 50 to 150 m2/m3 [15.2 to45.6 ft2/ft3] can be achieved. In contrast, plate-fin exchangers (right ) are typically used in cascadeliquefaction. A plate-fin exchanger uses layers of corrugated sheets or fins separated by metal plates.Hot and cold streams flow through alternating layers, and heat is transferred from the fin of one layerto the separating plate and then to the set of fins in the next layer—and finally to the other fluid.These exchangers are made as an all-brazed and welded pressure vessel with no mechanical joints.Similar to spiral-wound exchangers, plate-fin units in LNG service are typically aluminum and are verycompact—surface/volume ratios of 300 to 1,000 m2/m3 [91.5 to 305 ft2/ft3] can be achieved.

Tube stream out

Tube stream in

Vesselstream out

Vesselstream in

Mandrel

Fins

Parting sheets

Stream A in

Stream B out


Summer 2008 57

first single-train baseload plant built in the prior30 years.39 That first train and all subsequenttrain designs at the Point Fortin plant use threepure-component refrigeration cycles—propane,ethylene and methane. Heat exchange in therefrigeration units is carried out by plate-fin

exchangers and large gas turbines drive thecompressors. Each LNG train at Point Fortin hasparallel pairs of gas turbines and compressors.This allows each train to continue to functioneven with the loss of an individual compressor or turbine.40

Following liquefaction, the LNG produced atPoint Fortin is sent to storage tanks to awaitshipment by marine transport.41 The marinejetties at Point Fortin can accommodate LNGvessels up to 145,000 m3 [912,000 bbl], asignificant increase compared with the firstmarine LNG shipment of 50 years ago.

30. “Liquefied Natural Gas,” http://www.geoilandgas.com/businesses/ge_oilandgas/en/downloads/liquified_natural_gas.pdf (accessed June 11, 2008).

31. LNG plants may also use air-fin and shell and tubeexchangers, which are common in the oil field and inrefining. These exchangers are not discussed in this article.

32. Markussen D: “All Heat Exchangers Are Not CreatedEqual,” The Process Engineer (September 2004),http://www.chart-ind.com/literature_library_forall.cfm?maincategory=5 (accessed July 26, 2008).Markussen D: “Hot Technology for Lower Cost LNG,”Hydrocarbon Engineering 10, no. 5 (May 2005): 19–22.Markussen D and Lewis L: “Brazed Aluminum Plate FinHeat Exchangers—Construction, Uses and Advantagesin Cryogenic Refrigeration Systems,” presented at theSpring Meeting of the American Institute of ChemicalEngineers, Atlanta, Georgia, USA, April 10−14, 2005.

33. “Looking Inside…Spiral Wound Versus Plate-Fin HeatExchangers,” http://www.linde-plantcomponents.com/documents/looking_inside_PFHE_SWHE.pdf (accessedJune 15, 2008).

34. Core-in-kettle is a specialized application of plate-fintechnology.

35. LNG storage tanks at the liquefaction plant are similar tothe storage tanks at import terminals. Technology usedin these specialized tanks will be covered in a followingsection on import terminals.

36. Shareholders of Atlantic LNG Company include BP,British Gas, Repsol, Suez LNG and the National GasCompany of Trinidad and Tobago.

37. Trains are parallel production lines for LNG. For moreinformation: Hunter P and Andress D: “Trinidad LNG—The Second Wave,” presented at Gastech 2002, Doha,Qatar, October 13−16, 2002.

Diocee TS, Hunter P, Eaton A and Avidan A: “AtlanticLNG Train 4, The World’s Largest LNG Train,” presentedat LNG 14, Doha, Qatar, March 21−24, 2004.

38. Davis and Gold, reference 14.39. Redding P and Richardson F: “The Trinidad LNG Project–

Back to the Future,” LNG Journal (November−December 1998), http://lnglicensing.conocophillips.com/publications/index.htm (accessed July 26, 2008).

40. Although shutdown of an individual compressor orturbine would significantly reduce LNG productioncapacity, the train would avoid heating to ambientconditions and continue to operate until repairs were made.

41. LNG storage tanks at the liquefaction location are similarto the storage tanks at import terminals. Technologyused in these specialized tanks will be discussed in asubsequent section on import terminals.

> Atlantic LNG Company gas supply. The primary natural gas fieldssoutheast of Trinidad are about 55 km [34 mi] offshore and 110 km [68 mi]from Point Fortin. A complex system of subsea lines brings this gasonshore. For example, three lines —of 122-cm [48-in.], 91-cm [36-in.] and76-cm [30-in.]—take the gas from the Cassia area to Galeota Point andBeachfield onshore. Two 61-cm [24-in.] lines take gas from the Dolphinarea via an intermediate point at Poui to points onshore. Several otherlines connecting fields at Osprey, Teak, Mahogany, Flamboyant, Pelicanand Kiskadee/Banyan to the offshore pipeline system (not shown) roundout the picture. Multiple overland gas lines connect the onshorereceiving facilities with the Point Fortin LNG plant. Gas arrives from thenorthern Hibiscus field through a 61-cm underwater line. The northernfields lie about 32 km [34 mi] offshore and more than 83 km [52 mi] fromPoint Fortin.

A T L A N T I C O C E A N

C A R I B B E A N S E A

TOBAGO

TRINIDAD

Port-of-Spain

PointFortin Galeota

Point

Poui

Cassia

DolphinBeachfield

Hibiscus

200 miles

0 20km

> Atlantic LNG Company shipments. Atlantic ships product to importterminals in the Caribbean, United States and Europe (top left ). LNGimport terminals in the Caribbean include locations in the DominicanRepublic and Puerto Rico. Receiving terminals in the United States arelocated at Everett, Massachusetts; Cove Point, Maryland; Elba Island,Georgia; Gulf Gateway, Louisiana (offshore); and Lake Charles, Louisiana.European LNG shipments are primarily sent to terminals in Spain atBilbao, Huelva and Cartagena. On occasion, Atlantic has also shippedLNG to the UK, Japan and Belgium. The quantity of LNG shipped byAtlantic from Point Fortin (bottom right ) started modestly in 1999 at slightlyunder 2 million tonUK/yr [2.2 million tonUS/yr]. As additional capacity wasadded, these shipments expanded rapidly to nearly 13 million tonUK/yr[14.3 million tonUS/yr] shipped in 2006.

Mill

ion

tons

UK

/yr

14

12

10

8

6

4

2

0

Year

1999

2000

2001

2002

2003

2004

2005

2006

OtherSpainPuerto RicoUSA


Natural Gas at SeaThe first marine shipment of LNG was carriedaboard the Methane Pioneer in 1959, which hada capacity of only 5,560 m3 [35,000 bbl].42 Thatfirst load of LNG originated in Lake Charles,Louisiana, and was destined for Canvey Island inthe UK. That shipment and those that followed itclearly demonstrated that significant quantitiesof LNG could be safely transported on marinevessels. Since that time, large LNG transportshave become regular visitors to the world’scoastlines. Transports have grown in number,sophistication—and size (above). The numberof new LNG vessels has increased rapidly. In1990, the worldwide fleet numbered 70, whilethe current fleet has 266 carriers with 126vessels on order.43

High growth rate and change are not new tothe 50-year old LNG shipping industry. LNGvessels need a lead time of about four years tobuild and have a high capital cost—about twicethat of a large crude-oil carrier. In the past, LNGvessels were built in accordance with long-termcontracts to carry LNG from a specific lique -faction plant to designated import terminals.With the emergence of high prices for natural gasand spot market volatility, that feature of LNGshipping is changing.44 Liquefaction plantcapacity not tied up by contract is shipped to thelocation with the highest spot prices.

LNG vessel design is guided by severalcriteria that result from the physical charac -teristics of the LNG itself.45 First, the low densityof LNG dictates a large double-hull vessel withwater ballast, a low draft and high freeboard.46

The double hull serves as a safety feature andallows space for the water ballast. Secondly, theextremely low temperature of LNG requires use of special alloys for tank construction. Depending on the tank type, aluminum, stainless-steel and nickel-steel alloys may be used. Next,the high degree of thermal cycling in the onboardstorage tanks requires special care in design ofthe supporting structure. Lastly, because thecarbon steel commonly used in the hull isvulnerable to the extremely low temperatures ofthe LNG, good thermal insulation is required. Insome tank designs, the insulation must also becapable of supporting the cargo weight.

Application of these criteria has resulted inseveral LNG cargo designs with two systems ingeneral use—independent tanks and membranetanks. Independent tanks—such as the MossLNG transport system—are self-supporting anddo not form part of the ship’s hull. Membranetanks—such as those developed by Gaztransportand Technigaz—are supported by the vessel’shull through insulation and use a thin metalmembrane for containment (next page).47

As the trend to larger vessels has accelerated,both containment system designs haveencountered limitations. For independentspherical tanks, the weight and the specializedfacilities needed to construct them have provedto be a problem in some cases. In addition, shipswith spherical tanks pay higher Suez Canal feesthan other LNG ship types.48 On the other hand,membrane systems are susceptible to damagefrom sloshing loads caused by the large free-surface area in the tanks—a situation thatworsens as the vessel size increases. Research onsloshing using tank mock-ups is helping definethe best tank shape to resist this phenomenon.Although both tank containment systems are inwide general use, the membrane system is beingspecified for most of the very large LNG carrierspresently being built.49

While the containment system on LNGvessels continues to be a central area ofresearch, new concepts for vessel propulsionsystems are emerging as the industry focuses onemissions as well as high energy prices.Traditionally, LNG vessels have used steam-turbine propulsion systems that allow easydisposal of cargo boil-off gas.50 The industry hasbegun moving toward dual-fuel diesel engineswith an efficiency of 38 to 40% compared withsteam turbines at 28%.51 Systems with on-boardreliquefaction, combined gas turbine and steamturbine and boil-off gas reinjection are also beingconsidered. All of the new concepts inpropulsion, cargo containment and vessel designare being driven by high delivery costs andenergy prices and the desire to reduce emissions.The fate of one LNG vessel—the Polar Eagle—demonstrates how quickly LNG shipping haschanged in the past 15 years.

The Polar Eagle was built in 1993 at the IHIAichi Works in Nagoya, Japan, by ConocoPhillipsand Marathon Oil Corporation.52 On commis -sioning, the Polar Eagle was 230 m [755 ft] long,had a width of 40 m [131 ft], a gross weight of60,032 tonUK [66,174 tonUS] and carried a crewof 40. Propulsion was by steam turbine poweredby boil-off gas and heavy fuel oil.

This vessel was designed to transport87,500 m3 [550,660 bbl] of LNG in self-supporting,prism-shaped membrane cargo tanks. Unlikeother membrane-type tanks, this tank design canwithstand severe sloshing. For the last 15 years,this vessel transported LNG from theConocoPhillips Kenai LNG liquefaction plant toutility customers in Japan. Despite the fact thatthe vessel was still quite serviceable, the

58 Oilfield Review

> Evolution of LNG carrier capacity. During the last several decades LNGmarine carriers have grown significantly in cargo capacity. Standard-sizevessels of the last quarter of the 20th century had an LNG capacity 25times that of the original Methane Pioneer, and the current ratio is morethan 40. These capacity increases have been driven by a need to reduceshipping costs and to achieve economies of scale in vessel construction.LNG vessel capacity has had three distinct periods. During the earliestperiod—1965 to 1975—LNG carriers had a range of sizes, but were allrelatively small (green outline). Then came a longer period when most ofthe vessels were about 125,000 m3 [787,000 bbl], with gradual growth insize starting in the late 1990s (red outline). Currently, LNG carrier capacityis going through another step change. New super-sized LNG vessels aslarge as 265,000 m3 [1,668,000 bbl] have been built for long-haul service (blue outline).

LNG

car

rier

cap

acit

y, t

hous

and

m3

100

150

200

250

50

0

Delivery year

1965 1975 1985 1995 2005


Summer 2008 59

combination of small size and steam-turbinepropulsion made it uncompetitive on long-haulroutes. Recently, the vessel has been purchasedby Teekay Ltd to help its customers developsmaller gas fields and associated markets.53

Although LNG is an efficient way to bringstranded gas to market, there are combinations ofmarket size, shipping distance and reserve size inwhich neither LNG nor pipelines are economic.For these markets, compressed natural gas(CNG) may be an alternative solution.54 CNGtechnology reduces the volume of the natural gasbut stops short of liquefaction—significantlyreducing costs. One such CNG technology hasbeen developed by Sea NG and is named theCoselle high-pressure gas storage module.55

Coselle technology uses coils of CNG-filled steel

42. Foss, reference 17.43. Greer MN, Richardson AJ and Standström RE: “Large

LNG Ships—The New Generation,” paper IPTC 10703,presented at the International Petroleum TechnologyConference, Doha, Qatar, November 21−23, 2005.Vedernikova O: “LNG Shipping,” http://www.lngship.net/userFiles/2008%20Norton%20Rose.pdf (accessed June 17, 2008).

44. Valsgård S and Kenich A: “All at Sea,” LNG Industry(Spring 2007): 100–104.

45. Ffooks RC and Montagu HE: “LNG Ocean Transportation:Experience and Prospects,” Cryogenics 7, issue 1–4(December 1967): 324−330.

46. Draft is the depth of water required for the ship to float,and freeboard is that part of the ship between the deckand the waterline.

47. Deybach F: “Membrane Technology for Offshore LNG,”paper OTC 15231, presented at the 2003 OffshoreTechnology Conference, Houston, May 5−8, 2003.Kvamsdal R: “Spherical Tank Supported by a VerticalSkirt,” US Patent No. 4,382,524 (May 10, 1983).

48. Suez Canal fees are proportional to the internal volumeof a vessel. Moss LNG carriers have a much higherproportion of unused volume than do membrane carriersand hence pay higher fees.

49. Dabouis B: “Getting Gas to the Consumer,” LNG Industry(Spring 2008): 28−32.

50. LNG stays in the liquid form on marine vessels by auto-refrigeration. In spite of the insulation, enough heat isconducted through the insulated tank wall to causeslight boiling of the LNG. The small amount of gasformed is called the boil-off gas.

51. These engines are powered by diesel fuel or boil-off gas.In addition to the advantage in choice of fuels, dieselengines also have lower NOx emissions. NOx is a genericterm for oxides of nitrogen produced by combustion. Formore information: Kidnay and Parrish, reference 1.

52. “87,500 m3 SPB LNG Carrier Polar Eagle,” http://www.ihi.co.jp/ihimu/images/seihin/pl12_1.pdf (accessed June 17, 2008).

53. “Teekay Builds on Its LNG Service Offering,”http://www.marinelink.com/Story/TeekayBuildsonitsLNGServiceOffering-210674.html (accessed May 9, 2008).

54. One study found that CNG was most suited for transportdistances less than 2,500 km [1,550 miles]. For moreinformation: Economides MJ, Kai S and Subero U:“Compressed Natural Gas (CNG): An Alternative toLiquid Natural Gas (LNG),” paper SPE 92047, presentedat the SPE Asia Pacific Oil and Gas Conference andExhibition, Jakarta, April 5–7, 2005.

55. Stenning D: “CNG Opens New Markets,” Fundamentalsof the Global LNG Industry. London: PetroleumEconomist (2007): 67–68.

> LNG marine containment systems. Although several different marine containment systems for LNGhave been developed, only two systems are in widespread use today. Membrane tanks—found in 50%of the active fleet—use large tanks with a thin metal membrane to contain the LNG (bottom).Membrane tanks are supported by insulation between the metal membrane and ship’s hull. The metalmembrane may be a 35% nickel-steel, controlled-expansion alloy or stainless steel and has a typicalthickness of 0.7 to 1.2 mm [0.028 to 0.047 in.] depending on the metal employed. The insulationbetween the metal membrane and hull usually consists of two layers—plywood boxes filled withperlite or polyurethane foam insulation separated by another metal membrane barrier. The Mosssystem—seen in 47% of the active fleet—uses independent, spherical aluminum tanks to contain theLNG (top). These tanks are supported by a steel skirt and are not part of the hull structure. Mosstanks have three layers—an inner aluminum layer followed by insulation and an outer steel shell. Apipe tower shields inlet and outlet LNG lines.

Steel cover

Steel support skirt

Water ballast

Insulation

Aluminum shellPipe tower

Insulation layers

Metalmembranes


pipe in stackable units on a short-haul transportvessel (below). These vessels operate in a ferry-type system to ensure continual gas delivery.Technology developments in the CNG area mayopen smaller gas markets currently underservedby conventional delivery methods.

End of the ChainThe final link in the LNG delivery chain is theimport terminal. These terminals off-load theLNG from the marine vessel and store it ininsulated tanks until it is ready for regasificationinto the local transmission system. Worldwide,there are 60 LNG import terminals in operation,with 22 others in the construction stage (nextpage, top).56 LNG import terminals can be foundon every continent except Antarctica. About 50%of the terminals are in the Asia-Pacific region.Next in concentration is Western Europe with

nearly 25% of the import terminals. Theremainder are scattered across the globe.Currently, the USA has six terminals, and thehistory of LNG operations there illustrates the cyclical nature of the business during the last 30 years.

Between 1971 and 1980, natural gascompanies built four LNG import terminals inthe USA—Lake Charles, Louisiana; Everett,Massachusetts; Elba Island, Georgia; and CovePoint, Maryland.57 Delivery volume peaked in1979 but thereafter LNG imports rapidlydeclined. The decline was due to two factors—price disputes with Algeria and deregulation ofnatural gas in the USA that led to more USproduction. The terminals at Cove Point and ElbaIsland were mothballed in 1980, and theremaining two terminals experienced low volumein the following years.

In 1999, the convergence of three factorsmade LNG imports into the USA attractive again.First, Atlantic’s LNG plant in Trinidad started up,thereby reducing transportation costs. Secondly,increased natural gas demand was accompaniedby rising prices. Lastly, environmental concernsled to increased use of natural gas in electricpower generation. As a result, LNG importterminals at Elba Island and Cove Pointrestarted in 2001 and 2003, respectively. TheCove Point terminal provides a prime example ofan LNG import facility.

The Cove Point LNG import terminal islocated on the Chesapeake Bay about 120 km[75 miles] south of Baltimore, Maryland. CovePoint has an LNG storage capacity equivalent to221 million m3 [7.8 Bcf] of natural gas and apipeline delivery capacity of 28.3 million m3/d[1.0 Bcf/d]. The terminal connects to threenatural gas pipeline systems—TranscontinentalGas, Columbia Transmission and the Dominiontransmission system.

Operations at the Cove Point terminal aretypical of most LNG import facilities (next page,bottom). LNG vessels arrive from variouslocations, including Trinidad, Nigeria, Norwayand Algeria. LNG is offloaded from the carriers ata platform in Chesapeake Bay, about 4.0 km [2.5miles] offshore. From this platform, the LNG ispumped in insulated piping through anunderground, concrete-lined tunnel to insulated,double-walled storage tanks at the onshoreterminal. As LNG is required for sale, it is pumpedfrom the storage tanks to vaporizers and then intothe gas transmission system. Safety and securityat Cove Point include US Coast Guard supervisionof LNG vessels as they sail through ChesapeakeBay toward the unloading platform. The CoastGuard requires a security zone around the shipand offshore unloading platform—even when novessel is present.

LNG safety has come under increased scrutinysince September 11, 2001. Safety hazards directlyresult from physical properties of the LNG itselfand the resulting gas when vaporized. Thesehazards are cryogenic temperatures, gas-dispersion characteristics and combustibility.Since the industry’s inception in the 1940s, onlyfive accidents have occurred in or aroundliquefaction plants—unfortunately two of thoseresulted in deaths.58 The most seriousliquefaction plant accident occurred in Skikda,Algeria, in January 2004, when a steam boilerexploded, triggering a larger gas-vapor cloudexplosion.59 In addition, there have been two

60 Oilfield Review

> CNG shipping. The heart of the Coselle transport system is a 16-km [9.9-mi] length of 15.2-cm [6-in.]conventional steel pipe coiled into a stackable carousel that can be filled with CNG (bottom). Severalof these carousels can be loaded on a CNG transport vessel (top). (Graphic courtesy of Sea NGCorporation.)

56. “Liquefied Natural Gas Worldwide,” http://www.energy.ca.gov/lng/international.html (accessed May 15, 2008).

57. http://www.dom.com/about/gas-transmission/covepoint/index.jsp (accessed July 23, 2008).

58. “Liquefied Natural Gas Safety,” http://www.energy.ca.gov/lng/safety.html (accessed June 20, 2008).

59. The accident at Skikda killed 27 and injured 56. For moreinformation: http://www.ferc.gov/industries/lng/safety/safety-record.asp (accessed June 20, 2008).

60. A fire started in the tank resulting in a pressure increasethat lifted the concrete dome. The resultant collapsekilled 37.

61. Hightower M, Gritzo L, Luketa-Hanlin A, Covan J,Tieszen S, Wellman G, Irwin M, Kaneshige Melof B,Morrow C and Raglan D: “Guidance on Risk Analysisand Safety Implications of a Large Liquefied Natural Gas(LNG) Spill Over Water,” http://www.ferc.gov/industries/lng/safety/reports/sandia-rep.asp (accessed June 13, 2008).

62. Reference 58.


Summer 2008 61

import terminal accidents that resulted in deaths.The most serious import terminal incidentoccurred in 1973 at Staten Island, New York, USA,when a roof collapsed on an empty storage tank.60

However, these few isolated incidents are incontrast to the remarkably good safety record of

marine LNG transport. In past 40 years, more than80,000 LNG loads have been delivered withoutmajor accidents or safety issues.61

Operators who handle LNG have always hadsafety programs in place, but those programshave taken on greater importance in the last few

years. In 2003 and 2004, at least six major studieswere released that addressed LNG safety andsecurity.62 In addition to covering overall LNGsafety, these studies specifically addressed spillsover water, import terminal sites and quanti -fication of risks. While the LNG industry is

> Global LNG import terminals. Worldwide, there are 60 existing import regasification terminals located either onshore or offshore in 18 countries (green).An additional 22 terminal projects are under construction (red).

Existing terminals

Terminals under construction

> Import terminal components. LNG tankers arrive by marine transport at import terminal unloading platforms onshore or offshore. If the docking facilitiesand associated unloading platform are offshore, the LNG from the vessel is pumped through undersea piping to insulated storage tanks onshore.Insulated steel tanks are commonly used for storage, and they can be configured as single containment, double containment or full containment (fullcontainment shown). These tanks sit on a concrete base and have a 9% nickel-steel alloy inner lining covering outer shells of carbon steel and concrete.The roof of the storage tank is concrete over a suspended deck. As natural gas is required for distribution, it is pumped to a vaporizer. Although LNGstorage tanks are well-insulated, some boil-off always occurs. Boil-off gas can either be reliquefied or sent to the distribution system (not shown).

Vaporizer Pump Tunnel

LNG storagetank

Unloadingplatform

LNG tankerSuspended deckConcrete shell

Carbon-steel outer tankPerlite insulationNickel-steel liningConcrete pad


geographically quite diverse, four safetyelements have emerged that seem to encapsulatecurrent practice. These are primary contain -ment, secondary containment, safeguard systemsand separation distance.63

Primary containment is application of suit -able materials and design to contain LNG.

Secondary containment ensures that if spillsoccur, they can be contained and isolated.Safeguard systems act to minimize releases andmitigate their effects. Leak detection is oneexample of a safeguard system. Separationdistances relate to safety zones around shippinglanes and land-based installations. These foursafety elements apply across the LNG value chain.

While LNG has attracted some vocal criticswith regard to safety, the industry’s recordspeaks for itself. The need to bring distantnatural gas to local markets ensures that thistechnology will continue to have a significantrole in the energy arena.

62 Oilfield Review

> Darwin LNG. The Darwin LNG plant is located at Wickham Point in northwest Australia (left ). Natural gas for theDarwin plant is supplied from the Bayu-Undan field located between Darwin and East Timor in international waters.Gas wells at Bayu-Undan are in about 80 m [262 ft] of water, and reserves are estimated at 96.3 billion m3 [3.4 Tcf] ofgas and 65.6 million m3 [413 million bbl] of condensate. Gas arrives at the Darwin LNG plant through a 66-cm [26-in.]subsea pipeline.

AUSTRALIA

2000 miles

0 200km

EAST TIMOR

TIMOR

MelvilleIsland

Darwin

Bayu-Undan

AUSTRALIA

63. Foss MM: “LNG Safety and Security,”http://www.beg.utexas.edu/energyecon/lng/documents/CEE_LNG_Safety_and_Security.pdf (accessed May 15, 2008).

64. Yates D and Schuppert C: “The Darwin LNG Project,”presented at LNG 14, Doha, Qatar, March 21−24, 2004.

65. Montgomery T: “Aeroderivative Gas Turbine ProvidesEfficient Power for LNG Processing,” Pipeline & GasJournal (October 2001): 54, 56−57.

66. Kurbanov Y: “Russia to Become Key Player in World LNGover Next 10 Years,” http://www.oilandgaseurasia.com/articles/p/75/article/638/ (accessed July 22, 2008).

67. Terry MC: “Floating Offshore LNG Liquefaction Facility—A Cost Effective Alternative,” paper OTC 2215, presentedat the 7th Annual Offshore Technology Conference,Houston, May 5–8, 1975.

Barden JK: “Offshore LNG Production and StorageSystems,” paper SPE 10428, presented at the SPE Offshore Southeast Asia Show, Singapore,February 9−12, 1982.Faber F, Bliault AE, Resweber LR and Jones PS: “FloatingLNG Solutions from the Drawing Board to Reality,” paperOTC 14100, presented at the 2002 Offshore TechnologyConference, Houston, May 6–9, 2002.Wagner JV and Cone RS: “Floating LNG Concepts,”Proceedings of the 83rd Annual Convention of the TulsaGas Processors Association, 2004.Gervois F, Daniel L, Jestin N and Kyriacou A: “FloatingLNG—A Look at Export and Import Terminals,” paperOTC 17547, presented at the 2005 Offshore TechnologyConference, Houston, May 2–5, 2005.

Foss MM: “Offshore LNG Receiving Terminals,”http://www.beg.utexas.edu/energyecon/lng/documents/CEE-offshore-LNG.pdf (accessed May 15, 2008).

68. “ExxonMobil to Build First Gravity-Based Terminal inItaly,” http://www.poten.com/%5Cattachments%5C052305.pdf (accessed June 22, 2008).Sen CT: “LNG Trade Slows; Projects Advance,”http://www.ogj.com/print_screen.cfm?ARTICLE_ID=231654 (accessed June 22, 2008).

69. Krauss C:”Global Demand Squeezing Natural GasSupply,” http://www.nytimes.com/2008/05/29/business/29gas.html (accessed May 29, 2008).


Summer 2008 63

A Look at the FutureAs the LNG industry looks to the future,opportunities abound for application of newtechnology. At the start of the LNG chain, newliquefaction plants continue to push efficiencyimprovements that drive down operating costs. Agood example is the new ConocoPhillipsOptimized Cascade liquefaction plant at Darwin,Northern Territory, Australia (previous page).Completed in late 2005, the plant shipped its firstload of LNG to Japan in early 2006. At the time ofits startup, the Darwin plant pioneered severalfirsts.64 It was the first plant in the LNG industryto use high-efficiency, aeroderivative gas turbinesfor refrigerant compressor drivers.65 Theseturbines use less fuel, produce more LNG andhave the added benefit of reduced atmosphericCO2 and NOx emissions. It was the first operationto use turbine exhaust to provide heat for severalprocess areas. Finally, loading and vapor lines atthe Darwin plant use vacuum-insulated pipeinstead of conventional insulation.

New liquefaction plants—such as Darwin—are increasingly being built in remote, harshenvironments to move stranded gas to distantconsumers. A prime example of this trend is thelarge LNG plant being built on Sakhalin Island on

the northern Pacific coast of the RussianFederation. This plant is nearing completion andwhen fully operational will account for 5 to 6% ofworld LNG output.66 Buyers in Japan, Korea,Mexico and the United States have signed long-term contracts for LNG from Sakhalin.

Liquefaction plants are not the only part ofthe LNG chain to see improvements—thesetrends are also present in shipping. Shippingimprovements include new tank designs,shallow-draft hulls, twin-propeller carriers andmore efficient propulsion systems. Finally, ice-breaking vessel forms are being considered tobring LNG from stranded gas in arcticenvironments to consumers. All of theseimprovements will enhance and expand thecurrent global LNG shipping network (above).

The last part of the LNG chain—importterminals—may experience the most significanttechnology changes. Large carriers with theirneed for deep channels and berthing facilities—plus safety concerns—make offshore LNGinstallations attractive. The concept of offshoreor floating LNG installations is not new—theyhave been proposed and discussed for more than30 years.67 These proposals not only encompassimport terminals but also cover the entire range

of LNG facilities from liquefaction toregasification. Offshore LNG concepts are finallybeing put into practice with the completion ofthe Porto de Levante import terminal. Thisterminal—located off the Italian coast—willtake LNG from liquefaction plants in Qatar.68

Other offshore projects are in various stages ofplanning, permitting and construction.

Perhaps the most profound change along theLNG chain will involve commercial innovation—not technology. The LNG industry is on the cuspof changing from traditional long-term contractsto an emerging commercial trading model. Theeffects of this change are wrenching, and someearly participants have experienced unantici -pated shifts in demand.69 Even with earlysetbacks in commercial trading ventures, thedriving forces for LNG growth remain in place—and its future seems assured for decades. —DA

> Major LNG shipping routes. LNG from stranded gas reserves in the Middle East, Africa and the Caribbean is shipped to large consumers in Asia, Europeand the USA. Shipments to consumers in Japan and South Korea make up 54% of total marine LNG shipments. Shipments from Qatar to India, Nigeria toSpain and Trinidad and Tobago to the USA account for another 13% of the marine LNG trade. The remaining 33% of LNG transport comprises country-to-country shipments, each less than 8.0 billion m3 [282 Tcf]. These shipments total 73.8 billion m3 [2,606 Tcf] (not shown).

2007 LNG exports, billion m3

2007 LNG imports, billion m3

12.8

8.3 8.3

10.9

8.2

8.6

17.7

18.1

16.1

10.8


64 Oilfield Review

Triaxial Induction—A New Angle for an Old Measurement

Barbara AndersonConsultantCambridge, Massachusetts, USA

Tom BarberRob LeveridgeSugar Land, Texas, USA

Rabi BastiaKamlesh Raj SaxenaAnil Kumar TyagiReliance Industries Limited Mumbai, India

Jean-Baptiste Clavaud Chevron Energy Technology CompanyHouston, Texas

Brian CoffinHighMount Exploration & Production LLCHouston, Texas

Madhumita DasUtkal UniversityBhubaneswar, Orissa, India

Ron HaydenHouston, Texas

Theodore KlimentosMumbai, India

Chanh Cao MinhLuanda, Angola

Stephen WilliamsStatoilHydroStavanger, Norway

For help in preparation of this article, thanks to Frank Shray,Lagos, Nigeria; and Badarinadh Vissapragada, Stavanger.AIT (Array Induction Imager Tool), ECS (Elemental CaptureSpectroscopy Sonde), ELANPlus, FMI (Fullbore FormationMicroImager), MR Scanner, OBMI (Oil-Base MicroImager),OBMI2 (Integrated Dual Oil-Base MicroImagers) and Rt Scanner are marks of Schlumberger.Excel is a mark of Microsoft Corporation.Westcott is a mark of Acme United Corporation.

A new induction resistivity tool provides 3D information about formations far from the

wellbore. It improves the accuracy of resistivity measurements in deviated wells and

in dipping beds, and can measure formation dip magnitude and direction without

having to make contact with the wellbore. The tool’s highly accurate triaxial

resistivity measurement means fewer missed opportunities and better understanding

of the reservoir.

Triaxial induction resistivity is rejuvenating anold measurement. Formation resistivity, thefundamental property log analysts use to evaluateoil and gas wells, was the first measurementacquired with wireline logging tools. As theequipment to provide resistivity measurementsevolved, induction resistivity logging became thestandard measurement technique for acquiringformation resistivity. However, the accuracy oftool response at high resistivities and in deviatedwells or dipping reservoirs was limited by thephysics of the measurement. A new toolovercomes many of the limitations of previousinduction logging techniques. This 3D triaxialinduction measurement enables petrophysiciststo better understand and evaluate the types ofreservoirs where, before the new technology,hydrocarbons could have easily beenunderestimated or overlooked.

The resistivity story began a century ago,when Conrad Schlumberger developed atechnique for measuring the resistivity of thesubsurface layers of the Earth. His experimentsdemonstrated a practical application withcommercial possibilities. The concept waspromising enough that he formed a business

venture to put the technique into practice.1 OnSeptember 5, 1927, with equipment designed andbuilt by Henri-Georges Doll, the first electricallogging experiment, a measurement of formationresistivity, was conducted in a well in thePechelbronn oil region, France’s only large oilfield (next page, bottom).2

The fledgling oil and gas industry adoptedthis electrode-based resistivity measurement,and, with modifications, used it to identifyhydrocarbon deposits. Porous, permeable zoneswith high resistivity indicated the potential foroil or gas; low resistivity suggested the presenceof salt water. Then, in the 1940s, Doll introducedthe principles of induction resistivity logging tothe industry.3 This technique acquired formationresistivity in wells without a conductive path,notably in oil-base mud, overcoming a majorlimitation of electrode-based measurements.

The process of measuring formationresistivity is not as simple as taking a directreading from a tool or a measurement from Point A to Point B; however, in the past half-century, great strides have been made inaccurately measuring this critical parameter.Because induction logging tools provide

1. Gruner Schlumberger A: The Schlumberger Adventure.New York City: Arco Publishing, Inc., 1982.

2. Oristaglio M and Dorozynski A: A Sixth Sense: The Lifeand Science of Henri-Georges Doll Oilfield Pioneer andInventor. Parsippany, New Jersey, USA: The HammerCompany, 2007.

3. Doll HG: “Introduction to Induction Logging and Application to Logging of Wells Drilled with Oil-BasedMuds,” Petroleum Transactions, AIME 1, no. 6 (June 1949): 148–162.

4. For more on induction tool response: Gianzero S andAnderson B: “A New Look at Skin Effect,” The Log Analyst 23, no. 1 (January–February 1982): 20–34.


Summer 2008 65

apparent formation resistivity by taking ameasurement from a large volume of materialbeyond the borehole, all the components withinthat sensed region influence the final reading.Some of these interactions can negatively impactthe quality and accuracy of the measuredresistivity value.4 This is especially true when thelayers are not perpendicular to the axis of the

tool, as is the case with dipping beds anddeviated wells. Because of the effects of adjacentconductive layers, the resistivity measured byinduction logging tools in dipping beds may beconsiderably lower than the true resistivity,resulting in an underestimate of the hydrocarbonin place. Heterogeneity between the subsurfacestrata, and even within individual layers, alsoaffects tool response.

To account for these and other effects, loganalysts first used manual corrections and laterdeveloped computer-based, forward-modelingand inversion techniques to more closelyapproximate the true formation resistivity.However, they could not resolve all theunknowns—particularly formation dip. Despitethese unresolved errors in the measurement, the

Rh

Rv

Rh

Rv

Z

X

z

x

y

Y

Transmitter

Receiver

> The first resistivity log. The first carottage électrique (electrical coring) from a well in France’s Pechelbronn oil field was recorded on September 5,1927. The equipment to provide this resistivity log was based on tools used for surface mapping. The log is scaled in ohm.m, as are modern resistivitylogs. The high-resistivity interval correlated with a known oil sand in a nearby well, validating the use of log data to evaluate wells.

High resistivity


industry has successfully discovered much of theworld’s hydrocarbon resources using inductionlogging tools. Unfortunately, some reservoirshave been overlooked or underestimatedbecause of the measurement limitations.

Another difficult formation property forinduction tools to contend with is electricalanisotropy—variations in properties that changewith the direction of the measurement.5

Anisotropy is prevalent in shales as well as in theparallel bedding planes of laminated sand-shalesequences. When the beds are thinner than thevertical resolution of the induction logging tool,the measurement becomes a weighted average ofthe properties of the individual layers,dominated by the elements with the lowestresistivities. This phenomenon may mask thepresence of hydrocarbons.

The effects of anisotropy on the inductionresistivity measurement have been known sincethe 1950s, but until recently there has been noway to resolve the horizontal and verticalcomponents.6 By taking a 3D measurement—inessence a tensor rather than a scalar approach—these types of ambiguities and errors can be fullyresolved. However, sensors with the ability tomeasure induction resistivity in three dimensionsin tensor form had been beyond the limits ofexisting hardware. Similarly, the processingrequired to model and invert the measurementwas extremely time-consuming, even when usingsupercomputers or distributed networks.7

Many of the limitations inherent in inductionlogging have now been overcome with the Rt Scanner triaxial induction service. Currentlyavailable computational-processing power hasbeen combined with a new tool design to createa step change in the evolution of inductionlogging. This new tool is solving problems andproviding the industry with answers to questionsthat have plagued log analysts and geologistsfrom the beginning of well logging.

Three primary applications of triaxial induc -tion tools are accurate resistivity measure mentsin dipping formations, identi fication andquantification of laminated pay intervals and anew structural dip measurement that requires nopad contact. This article describes how thesemeasurements are made and demonstrates theirapplications. Also included are case studies fromAfrica, India and North America.

Induction Resistivity BasicsA two-coil array demonstrates the physics of atraditional uniaxial induction resistivity measure -ment. Alternating current excites a transmittercoil, which then creates an alternating-electromagnetic field in the formation (left).8

This field causes eddy currents to flow in acircular path around the tool. The ground loops ofcurrent are perpendicular to the axis of the tooland concentric with the borehole. They are atleast 90° out of phase with the transmittercurrent, and their magnitude and phase dependon the formation’s conductivity.

The current flowing in the ground loopgenerates its own electromagnetic field, whichthen induces an alternating voltage in thereceiver coil. The received voltage is at least 90°out of phase with the ground loop and more than180° out of phase with the transmitter current.Induction resistivity from the formation is derivedfrom this voltage, referred to as the R-signal.Direct coupling of the tool’s primary transmitter

66 Oilfield Review

> The concept of induction resistivity. The basic physics of the inductionresistivity measurement is represented by a two-coil array. A continuousdistribution of currents, generated by the alternating-electromagnetic fieldof the transmitter (T), flows in the formation beyond the borehole. Theseground loops of current generate electromagnetic fields that are sensed bythe receiver coil (R). A phase-sensitive detector circuit, developed originallyfor land-mine detection during World War II, separates the formation signal(R-signal) from the directly coupled signal coming from the transmitter (X-signal). The R-signal is converted to conductivity, which is then convertedto resistivity. (Adapted with permission from Doll, reference 3.)


Summer 2008 67

field in the receiver coil, the X-signal, combineswith the formation R-signal; however, the directlycoupled signal is out of phase with thecontribution from the formation. This phasedifference, detected using phase-sensitivecircuitry, permits the rejection of the X-signal andmeasurement of the R-signal.

Conversion of the R-signal voltage toconductivity was first accomplished by equationsbased on the Biot-Savart law, which assumes themajor contribution of a single ground loop willhave a maximum value at the midpoint of thetransmitter and receiver coils.9 Schlumbergermathematicians later developed equations—based on the complete solution for Maxwell’sequations—that provided more accurate measure -ments.10 This solution can be visualized using asimplified version of Maxwell’s equations—theBorn approximation—which is an acceptedmethod of determining the source and location of the formation signal. For the two-coil axial array, the response is essentially a toroid shapesurrounding the tool and perpendicular to its axis,with maximum values near the midpoint of thetransmitter and receiver (right).11

In vertical wells with thick homogeneoushorizontal beds, standard resistivity loggingtools, such as the AIT Array Induction ImagerTool, work reasonably well. These uniaxial toolsmeasure apparent resistivity, Ra, in a horizontalplane, which is equivalent to horizontallymeasured resistivity, Rh. Resistivity measured ina vertical plane, Rv, cannot be measured withuniaxial induction tools in a vertical well.

Because the ground loops of induction toolsintersect a huge volume of the formation, theymay traverse a path that includes several differentlayers with varying electrical proper ties.Anisotropy results in a resistivity measurementthat changes based on the direction of themeasurement. This limitation in the measurementwas one of the factors that led to the developmentof the Rt Scanner tool.

The Impetus for Triaxial MeasurementsAlthough the concepts underlying triaxialinduction measurements first appeared in theliterature in the mid 1960s, the tools to make thismeasurement were not developed. There werethree main reasons for the delay: a triaxial toolcould not be built with the existing technology,the data processing required was beyond thecapability available at the time, and the tool’sresponse to conductive fluids in the borehole could be much larger than the signal from the formation.

Interest in triaxial induction was renewedchiefly because of the recognized limitations ofuniaxial resistivity measurements in two areas:anisotropic reservoirs and bedding planes thatare not perpendicular to the axis of the tool.12

Although both of these limitations wereidentified in the 1950s, there was then no direct

method of measuring anisotropy with aninduction logging tool, and the solution tonegative effects of real or relative dipping bedson induction resistivity was not trivial.13 Astechnology advanced, measurement under -standing, processing power and tool design allplayed key roles in solving for these effects,

5. For more on anisotropy: Anderson B, Bryant I, Lüling M,Spies B and Helbig K: “Oilfield Anisotropy: Its Origins andElectrical Characteristics,” Oilfield Review 6, no. 4 (October 1994): 48–56.Tittman J: “Formation Anisotropy: Reckoning with ItsEffects,” Oilfield Review 2, no. 1 (January 1990): 16–23.

6. Kunz KS and Gianzero S: “Some Effects of FormationAnisotropy on Resistivity Measurements in Boreholes,”Geophysics 23, no. 4 (October 1958): 770–794. Moran JH and Gianzero S: “Effects of FormationAnisotropy on Resistivity-Logging Measurements,” Geophysics 44, no. 7 (July 1979): 1266–1286.

7. Anderson B, Druskin V, Habashy T, Lee P, Lüling M, Barber T, Grove G, Lovell J, Rosthal R, Tabanou J,Kennedy D and Shen L: “New Dimensions in ModelingResistivity,” Oilfield Review 9, no. 1 (Spring 1997): 40–56.

8. For a detailed explanation of induction theory: Moran JHand Kunz KS: “Basic Theory of Induction Logging andApplication to Study of Two-Coil Sondes,” Geophysics 27,no. 6, part I (December 1962): 829–858.

9. The Biot-Savart law describes the magnetic field generated by an electric current.

10. Maxwell’s equations, named for physicist James ClerkMaxwell, are a set of four partial differential equationsthat explain the fundamentals of electric and magneticfield relationships.

11. Habashy T and Anderson B: “Reconciling Differences inDepth of Investigation Between 2-MHz Phase Shift andAttenuation Resistivity Measurements,” Transactions ofthe SPWLA 32nd Annual Logging Symposium, Midland,Texas, June 16–19, 1991, paper E.

12. Moran and Gianzero, reference 6.13. For the theoretical solution to Maxwell’s equations as

applied to induction logging: Moran and Kunz,reference 8. Anderson B, Safinya KA and Habashy T: “Effects of Dipping Beds on the Response of Induction Tools,” paper SPE 15488, presented at the SPE Annual Technical Conference and Exhibition, New Orleans,October 5–8, 1986.

> Born approximation for a uniaxial induction logging tool. The sensed regionfor uniaxial induction tools is a toroid shape (red), perpendicular to the tool.The maxima are located approximately at the midpoint between the transmitter(T) and receiver (R). This rendering shows the Born approximation of the fullsolution to Maxwell’s equations. The shape is valid for thick beds andhomogeneous, isotropic formations. This region sampled by the uniaxialinduction tool corresponds to only one of the nine modes measured by thetriaxial Rt Scanner tool.

T

R


ultimately resulting in the development of atriaxial induction tool (below left).

Developing such a tool involved understandingthe effects of the borehole on the measurement.14

There is a great sensitivity to eccentricity in theborehole: the more conductive the mud, thegreater the effect. The sensitivity results in theformation signal being overwhelmed by theborehole signal. This situation, the effects ofwhich can be two orders of magnitude greater fortriaxial tools than for uniaxial induction tools,would have been an insurmountable obstaclewithout intensive computer modeling.

Iterative modeling allowed various triaxialtool designs to be tested without having to buildand test physical tools. Final tool design includeda sleeve with electrodes connected to aconductive copper mandrel. This configurationreturned the borehole currents through the tool,reducing the large signals caused by thetransverse eccentricity to a level equivalent tothat of the AIT tool. The correction for boreholeeffects could then be handled in a mannersimilar to that used for the AIT measurement.15

After engineers solved for borehole effects,tool response to various geometrical scenarioswas investigated. For most of their history,induction measurements have had to contendwith geometry, both in the borehole and in the

formation. Geometry was regarded by inter -preters as a major nuisance or, at best, somethingto be coped with.16 However, after the AIT tool’sresponse was modeled, tool designers discoveredthat the formation-geometry effects are thestrongest contributor to the induction signal.When properly resolved and modeled, geometrynow provided a key to accurate measurement offormation resistivity. In addition, dipping beds—those that are not perpendicular to the axis ofthe logging tool—could be properly measured.

Dipping beds are the result of geologicaltilting of formations, deviation of the wellboretrajectory from vertical, or combinations of both.Fast analytical codes, developed in the 1980s,estimate resistivity in dipping beds using datafrom uniaxial induction tools, but the processing

68 Oilfield Review

> Rt Scanner triaxial induction service. The RtScanner tool comprises a triaxial transmitter,three short-spacing axial receivers for boreholecorrections and six triaxial receivers. Electrodeson the tool and the Rm sensor in the bottom nose,which measures the mud resistivity, are alsoused for borehole corrections. An internal metalmandrel (not visible in the drawing) provides aconductive path for borehole currents to returnthrough the electrodes on the exterior of the tool.

Electronics housing

Triaxial transmitter

Three short uniaxialreceivers for boreholecorrection

Six triaxial receivers

Metal mandrel

Sleeve with shortelectrodes

Rm sensor

Triaxial transmitter

Triaxial receiver

Axial receiver

Electrode > Three-dimensional arrays. The Rt Scanner service produces a nine-elementarray for each transmitter and receiver pair. Traditional induction measurementsare made by passing current through coils that are wrapped around the axisof the tool, also called the z-axis (blue), which induces current to flow in theformation concentrically around the tool. Triaxial induction tools also includecoils that are wrapped around the x-axis (red) and y-axis (green), whichcreate currents that flow in planes along the tool’s x- and y-axes. The x, y andz components of the transmitter couple with the x, y and z receivers. Forvertical wells with horizontal beds, only the xx, yy and zz couplings respond tothe conductivity (σ) of the formation. In deviated wells or wells with dippingbeds, all nine components of the array are needed to fully resolve theresistivity measurement. The multiple triaxial transmitter and receiver pairsgenerate 234 conductivity measurements for each depth frame.

Tz

Rz

Tx

Rx

Ry

Ty

xx xy xz

yx yy yz

zx zy zz

=


Summer 2008 69

relies on inputs from other sources.17

Unfortunately, the uniaxial measurement maybecome unreliable or provide nonuniquesolutions when external data sources are used.

All these issues posed problems for uniaxialinduction tools. In most cases, there was notenough information to fully correct the data.Triaxial induction tools, however, make thenecessary measurements to resolve the ambi -guities and properly measure the resistivity ofanisotropic reservoirs, correct for nonuniformfiltrate invasion, correct for the effects of dipping beds and deal with geometrical effectson the measurement.18

Triaxial Resistivity Theory Previous induction logging tools, such as thosefrom the AIT family, measure horizontalresistivity (uniaxially). The Rt Scanner toolmeasures in three dimensions (triaxially).Although the physics of measurement aresimilar, triaxial tools are much more complex(previous page, bottom right).

The Rt Scanner tool consists of a collocatedtriaxial transmitter array, three short axial

receivers and three collocated triaxial receiverarrays. The triaxial transmitter coil generatesthree directional magnetic moments in the x, yand z directions. Each triaxial receiver array hasa directly coupled term and two terms cross-coupled with the transmitter coils in the otherdirections. This arrangement provides nineterms in a 3x3 voltage tensor array for any givenmeasurement. All nine couplings are measuredsimultaneously. An advanced inversiontechnique extracts resistivity anisotropy, bed-boundary positions and relative dip from thetensor voltage matrix. The receiver arrays arelocated at different spacings to provide multipledepths of investigation.

The Born approximation for the triaxialinduction tool’s response provides a graphicalrepresentation for the solution of the equationsrepresenting the sensed region (above). Theuniaxial induction tool’s response was shownearlier to have a single toroid shape; the triaxialtool delivers nine responses superimposed oneach other. The zz term from the Rt Scanner toolis essentially the same response as thatmeasured by the uniaxial induction tool.

Collocation of the coils is an importantfeature of the Rt Scanner tool: when thetransmitter or receivers are not at the sameposition, the spacings for the cross-terms will bedifferent from those of the direct terms. Becausethe entire ensemble of measurements is madewithin a single depth frame, no measurements

14. Rosthal R, Barber T, Bonner S, Chen K-C, Davydycheva S,Hazen G, Homan D, Kibbe C, Minerbo G, Schlein R, Villegas L, Wang H and Zhou F: “Field Test Results of anExperimental Fully-Triaxial Induction Tool,” Transactionsof the SPWLA 17th Annual Logging Symposium, Galveston, Texas, June 22–25, 2003, paper QQ.

15. For details on Rt Scanner design and modeling: Barber T, Anderson B, Abubakar A, Broussard T, Chen K-C, Davydycheva S, Druskin V, Habashy T, Homan D, Minerbo G, Rosthal R, Schlein R and Wang H:“Determining Formation Resistivity Anisotropy in thePresence of Invasion,” paper SPE 90526, presented atthe SPE Annual Technical Conference and Exhibition,Houston, September 26–29, 2004.

16. Moran and Gianzero, reference 6.17. Barber TD, Broussard T, Minerbo G, Sijercic Z and

Murgatroyd D: “Interpretation of Multiarray Logs inInvaded Formations at High Relative Dip Angles,” TheLog Analyst 40, no. 3 (May–June 1999): 202–217.

18. During the drilling process, fluids from the drilling mudleave the wellbore and enter permeable formations. Themud filtrate alters the electrical characteristics of theformation around the wellbore. The depth of filtrate inva-sion, and its associated geometry, may be unpredictable.

> Born approximation for a triaxial induction tensor voltage array. The Born response function for a triaxial induction tool ismuch more complex than that for a uniaxial induction tool. There are nine elements, one for each component of the tensorvoltage array. Each transmitter-receiver pair has positive (red) and negative (blue) responses. The surfaces represent theregions where 90% of the signal measured by the receiver coil originates. Each of the nine components is superimposed atthe measure point of the tool. The xx, yy and zz elements are derived from the direct coupling of a triaxial transmitter and itsassociated triaxial receiver. The other six elements represent cross-coil responses. The zz response (bottom right ) is theonly one measured by the simpler uniaxial induction tool.

50

–50

0

z-ax

isy-axis

xx

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

yx

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

zx

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

xy

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

yy

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0z-

axis

y-axis

zy

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

xz

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

yz

x-axis

10050

0–50

–100 –100–50

050

100

50

–50

0

z-ax

is

y-axis

zz

x-axis

10050

0–50

–100 –100–50

050

100


have to be depth-shifted to form the measure -ment tensors. When all nine components are atthe same spacing and location, the matrix can bemathematically rotated to solve for relativeformation dip. A change from one coordinatesystem to another is also greatly simplifiedbecause it involves a simple transformation, andall measurements are made along the samecoordinate system as well as at the same depth.Collocation is especially important when beddingplanes are not perpendicular to the relativeposition of the tool.

Power in the ProcessingCollocated orthogonal transmitter and receiverpairs made the triaxial resistivity measurementfeasible, but advancement in processing powerwas the enabler that spurred the development ofthe tool. Even in the late 1990s, triaxial inductionwas referred to as a theoretical concept, prima -r ily because the computing power needed tomodel and develop fast processing codes was notreadily available.19 Moore’s law, the observationthat computing power doubles every two years, isevidenced in the progression that has occurredwith induction resistivity logging.

The first induction resistivity tools convertedconductivity measured downhole to an analogvoltage that was measured at the surface. The loganalyst read the resistivity from the logs andapplied corrections from charts to account forthe effects of adjacent beds and filtrate invasion,generally ignoring borehole effects. Boreholecorrection charts were then developed based ongeometrical-factor curves obtained from labora -tory measurements made in plastic pipesimmersed in waters of varying salinity.20 In themid 1980s, these empirically derived charts werereproduced using computer modeling.

70 Oilfield Review

0–2,500 –2,000

1 10 100

–1,500 –1,000 500

Conductivity, mS/m

Resistivity, ohm.m

Conductivity, mS/mConductivity, mS/m

0 500 1,000 1,500 2,000 –2,500 –2,000 –1,500 –1,000 500 0 500 1,000 1,500 2,000 –2,500 –2,000 –1,500 –1,000 500 0 500 1,000 1,500 2,000

10 xxxyxzyxyyyzzxzyzzσhσv

20

30

40

Dep

th, f

t

50

60

70

80

0

10

20

30

40

Dep

th, f

t

50

60

70

80

Rh

Rv

Rh (inverted)

Rv (inverted)

80 ft

50 ft

40 ft

30 ft

20 ft

0 ft

Rh = 1.9 ohm.mRv = 11.0 ohm.m

Rh = 1 ohm.mRv = 2 ohm.m

Rh = Rv = 50 ohm.m

Rh = Rv = 0.5 ohm.m

Rh = Rv = 1 ohm.m

Θ

Φ

> Modeling the triaxial induction response. A 1D horizontally layered,transversely isotropic (TI) model was used to validate the triaxial inductionresponse to known conditions (bottom right ). The five layers used in themodel consist of two low-resistivity homogeneous layers, a high-resistivityhomogeneous layer, and two anisotropic layers with high- and low-contrast beds. The first measurement is conducted with a vertical tool inhorizontal beds (top left ). The zz (blue) and yy (green) components react tothe resistivity of the beds, but the xx and all cross-components are zero.Prior to inversion, none of the curves indicates the correct horizontal (pinkdash) and vertical (black dash) conductivity. Next, the model well isdeviated 75° (Θ) and the tool position is rotated 30° (Φ) from the high sideof the wellbore. All nine components become active (center) and nonereads the same as the vertical model. The zz (blue) componentcorresponds to a uniaxial induction measurement, and although it is similarto the curve in the vertical response model, the curve’s shape andamplitude have changed. The data are then rotated mathematically (topright ) to zero the yx and yz (green dash) cross-coil contributions. Theangle of rotation required to zero these components corresponds to therelative dip of the beds. Finally, the data are inverted, correcting for bedthickness and deviation, and converted from conductivity to resistivity(bottom left ). In the three lower layers, which are homogeneous, Rv (blue)and Rh (red) are equal and match the input resistivity. In the laminatedlayers, the curves separate as a result of anisotropy.


Summer 2008 71

The manual process of correcting inductionlog data was carried out sequentially: applyborehole corrections, correct for shoulder-bedeffects and correct for invasion. With the advent ofdata recorders, log data could be processed usingcomputers. Codes were developed to perform 1Dcorrections automatically, first at mainframe-equipped computing centers and then asprocessing power continued to grow, at thewellsite using computer-equipped logging units.

Advances in computer technology renderedthe manual corrections obsolete, but there was aproblem in the methodology. The codes weredeveloped assuming horizontal, homogeneousbeds, and corrections were applied with thesame linear approach used by log analysts.However, the ground loops produced by inductiontools intersect and interact with all the mediathey come into contact with in a complex,nonlinear fashion.21 The sequential approach,used for decades, was found to be inadequate.

This situation was improved when fast 2Dasymmetric forward-modeling codes weredeveloped in the mid 1980s. They revealed justhow inaccurate sequential chartbook correc tionswere for determining the true resistivity, Rt,especially in thin beds invaded by mud filtrate.Development of the AIT tool was a result oflessons learned from those models. Since then,various techniques have been applied to obtainRt, including iterative forward modeling andinversion.22 Models have been developed thatinclude 1D corrections as well as corrections forinvasion and nonhorizontal bedding (2D) and

nonlinear invasion in tilted reservoirs (3D). Onlyrecently has advanced computer-processingpower enabled inversion codes that fully correctthe induction measurement. These codes allowsimulations to be run in hours instead of weeks.If Moore’s law holds true, hours for processinginduction measurements will eventually bereduced to seconds.

Induction resistivity data, acquired with atriaxial tool, could now be processed in areasonable time frame. All the pieces of thepuzzle were available; the next step was to putthe triaxial tool to the test.

Testing the CodeTo test the validity of the acquisition andinversion algorithm for triaxial induction data, a1D horizontally layered, transversely isotropic(TI) model was constructed (previous page).Five layers simulated a complex reservoircomprising two low-resistivity sands, a high-resistivity sand, an anisotropic low-resistivityshale and a laminated sand-shale sequence.

This simulated reservoir included featuresthat present limitations for uniaxial resistivitytools. The testing proved that a triaxial resistivity measurement overcomes theselimitations and provides accurate resistivity inchallenging environments.

The outputs of the processing are trueresistivity corrected for dip in the nonlaminatedlayers and a shale-affected resistivity inlaminated layers. Rv is provided from theprocessing, although it is equivalent to Rh in theisotropic intervals.

For the two laminated layers, Rv and Rh arenot equal, and the curves have separation basedon the degree of anisotropy. Neither Rh nor Rv

provides the true resistivity of the modeledreservoir in the case of laminated sections, buttechniques have been developed to provide theresistivity of the sand layers.

True ResistivityThe true resistivity of a formation, Rt, is acharacteristic of an undisturbed, or virgin,region. Much study and research have beencarried out in the name of acquiring this elusivemeasurement. The measurement of inductionresistivity in a virgin zone is predicated on somedegree of homogeneity, consistent perpendicularbeds and isotropic reservoirs. In nature, this israrely the case.

The concept of vertical and horizontalresistivities evolved early in the development ofelectrical logging. Measured apparent resis tivity,Ra, of stacked rock layers differs with changes inthe measurement direction. If the measurementis made parallel to the layers, the result issimilar to measuring resistors in parallel—thelowest resistances dominate (above). For aparallel resistor circuit, more current flowsthrough the smaller resistors, and each resistor

19. Anderson BI: Modeling and Inversion Methods for theInterpretation of Resistivity Logging Tool Response. Delft,The Netherlands: Delft University Press, 2001.

20. Moran and Kunz, reference 8. 21. Anderson, reference 19.22. Howard AQ: “A New Invasion Model for Resistivity Log

Interpretation,” The Log Analyst 33, no. 2 (March–April 1992): 96–110.

> Direction matters. Under the right conditions, the deep-induction response to a homogeneous, isotropic bed (left ) is the same as that to an anisotropic,laminated bed (center). This occurs when beds are thinner than the vertical resolution of the measurement. For the 90-in. deep-induction array, thevertical resolution is 1 to 4 ft [0.3 to 1.2 m]. Horizontal resistivity (Rh) measurements are analogous to parallel resistor circuits, so the resistivity value of thelaminated bed is primarily influenced by the layer with the lowest resistivity, Rshale. With standard induction tools, hydrocarbon-bearing sand layers caneasily be overlooked. Vertical resistivity (Rv) is analogous to a series resistor circuit (right ), and its value is dominated by the layer with the highestresistivity. A large difference between Rv and Rh indicates anisotropy.

1,800

Depthft

Computed Deep Induction

ohm.m0.2 2,000

1,810

1,820

1,830

1,840

1,800

Depthft

1,810

1,820

1,830

1,840

Computed Deep Induction

Model Rt Profile Model Rt Profile Model Rh-Rv Profile

Rh Rv

ohm.m0.2 2,000

Horizontal Resistivity, Rh

Vertical Resistivity, Rv

ohm.m0.2 2,000R

sandR

shale

Rsand

Rshale

Rshale

Rsand

Rsand


divides the current according to the reciprocalof its resistance.

When the measurement is made across thestack, the measured resistance is similar tomeasuring resistors in series. In an electricalseries circuit, the resistance values are addedtogether. Higher resistance, which is the case forthe layers containing hydrocarbon, is dominant.

The concept that the measured resistancedepends on the direction in which it is made isreferred to as electrical anisotropy. Since welllogging began in vertical wells with stacks ofmore or less horizontal layers, the resistivityparallel to the layers was called the horizontalresistivity, Rh, and the resistivity measuredacross the layers was called the verticalresistivity, or Rv. In an isotropic, thick sand Rh =Ra = Rv. If, however, the thickness of the beddinglayers is less than the tool’s vertical resolution,the Rh measurement is analogous to the parallelelectrical circuit.

Most of the technology for determiningformation resistivity measured the horizontalcomponent, giving rise to difficulties inevaluating thin layers comprising shale andhydrocarbon-bearing sands. For a uniaxialinduction measurement the formation currentsflow in horizontal loops, and the resultingsensitivity is to the horizontal resistivity. Formost laminated reservoirs, Rh ≠ Rv. Based on theparallel circuit analogy, Ra will be similar invalue to that of the layer with the lowerresistivity, usually the shale. Therein lies theproblem with interpreting induction resistivity inlaminated reservoirs: the dominant nature of theless-resistive layers masks the more-resistivelayers that may have hydrocarbon potential. Theresult is that pay zones may be overlooked orunderestimated.23 The Rv /Rh ratio is a usefulmeasurement for determining the level ofanisotropy, and when the ratio is higher than 5, it alerts the log analyst to look for potentiallaminated-pay reservoirs.

For a laminated sand-shale sequence, theportion of the reservoir that is of interest is thesand. Although Rv does not provide the actualresistivity of the hydrocarbon-bearing sand layer,Rsand, it can be combined with othermeasurements to derive it. The shale effectsmust be removed from the volumetricmeasurement to obtain the resistivity of the sandlayers (above). Calculating Rsand from Rh and Rv

requires a secondary source to determine thevolume of shale before its effects can beeliminated. Shale volume can be obtained fromseveral sources, including the ECS ElementalCapture Spectroscopy sonde. Once determined,Rsand can be used to calculate water saturation,Sw, using Archie’s equation. The full derivation ofthe formula for Rsand and Sw in the presence ofanisotropy can be found in the literature.24

72 Oilfield Review

> Hidden saturation. Rh and Rv are outputs from the Rt Scanner tool. The resistivity of the sand layers can beresolved from these measurements in combination with fractional volumes of sand and shale. For this example,the conventional induction tool would have measured Rh = 2.3 ohm.m. Rv from the triaxial induction measurementis 12.8 ohm.m. The volume fractions, Fshale and Fsand, could come from an ECS Elemental Capture Spectroscopytool. Because shales often exhibit anisotropy without the presence of sand laminations, two different shalevalues are used in this example: vertical Rshale-v is 2 ohm.m and horizontal Rshale-h is 1 ohm.m. These values shouldbe determined within an anisotropic shale interval. This method gives an Rv /Rh ratio in the shale of 2, comparedwith the 5.6 ratio of the entire sand-shale sequence. Solving the equations (right ) for Rsand yields a value of 20 ohm.m.The 2.3 ohm.m measured by a conventional induction tool would considerably underestimate the hydrocarbon volume.

Rsand

Rsand

Rsand

Rshale-h

Rsand

Rshale-h

Rshale-h

Rshale-v

Rshale-v

Rshale-v

Rsand

Rsand

Rshale–h = 1 ohm.m

Rshale–v = 2 ohm.m

Rv = 12.8 ohm.m

Rh = 2.3 ohm.m

1

Rh

= +Fsand

Rsand

Fshale

Rshale-h

Rv = +x xFsand Rsand

Fshale = 40%

Fsand = 60%

Rsand = 20 ohm.m

Fshale Rshale-v


Summer 2008 73

The calculation of Rsand and Sw in the sandfraction is typically carried out usingpetrophysical analysis software. However, Excelspreadsheets have been developed to manuallyconvert Rv and Rh to water saturation, Sw.25

The two major limitations of uniaxialinduction tools, incorrect resistivity in dippingbeds and anisotropy effects, are overcome by thetriaxial induction measurement. More accurateresistivity leads to more accurate Sw, whichenables petrophysicists to correctly evaluatehydrocarbon reservoirs. Properly characterizinglaminated sands means fewer missed low-resistivity reservoirs. True resistivity in deviatedwells and dipping beds means more accuratevolumetric analysis. Ultimately, more oil and gascan be discovered and produced from reservoirs.The following case studies demonstrate howtriaxial resistivity measurements are used toevaluate difficult-to-interpret oil and gas wells.

True Resistivity in Deviated WellsAn AIT tool was run offshore Angola in a well thatwas deviated 60°. The formations encounteredincluded two 30-ft [10-m] sands with highresistivity. A 30-ft interval is generally within thevertical resolution of this uniaxial tool andtherefore should provide a reasonable Rt readingfrom the deepest induction measurement, the90-in. array. However, because of effects of welldeviation on the measurement, the 90-in.resistivity reading was lower than the actual Rt.

An Rt Scanner tool was then run over thesame interval. Inverting the data and correctingfor the effects of dip produced resistivity valuesthat were more accurate than those of the AITtool (above right). The corrected resistivity fromthe Rt Scanner tool is five times greater than thedeep resistivity value from the AIT tool.

Although the water saturation calculatedwith the resistivity from either tool wouldindicate the presence of hydrocarbons, reservecalculations would be substantially different.Higher hydrocarbon saturations and volumescalculated using outputs from the Rt Scannertool would affect production-facility design, long-term infrastructure planning and recoverabilitydecisions in the event of secondary and tertiaryrecovery programs. Having an accurate Rt valuehas enormous implications, especially formarginal reservoirs, where critical go/no-godecisions based on less-accurate data wouldunderestimate hydrocarbon in place.

An additional consideration is that the cost ofdrilling deepwater prospects has limited thenumber of wells that can be drilled to evaluate aprospective reservoir. Petrophysicists and

geologists must construct reservoir models withsurface-acquired data validated by fewer actualwells. It is absolutely crucial that these modelsbe calibrated to the most accurate dataavailable, because the luxury of drilling step-outand infield wells to refine the models isprohibitively expensive. It is more cost-effectiveto use accurate triaxial induction resistivity datacorrected for dip and deviation to improvereservoir understanding with the very first well.

Anisotropy in Deepwater Turbidites E&P companies cannot afford to underestimatereserves or miss opportunities. Unfortunately,laminated sand-shale sequences have been

overlooked because of the effects of anisotropy.Examples of laminated reservoirs are turbiditeand fluvial deltaic sediments. The term “low-resistivity pay” has been applied to these types of environments.

Anisotropy-related suppression of theresistivity values measured by traditionalinduction logging tools is the predominantreason for the low resistivity. But even whenthese reservoirs are correctly identified, they aredifficult to evaluate. In practical terms, usingconventional resistivity measurements to calcu -late hydrocarbon reserves may result inunderestimates of more than 60% compared withanalysis using Rv and Rh.26 The Krishna-Godavari

23. Boyd A, Darling H, Tabanou J, Davis B, Lyon B, Flaum C,Klein J, Sneider RM, Sibbit A and Singer J: “The Lowdown on Low-Resistivity Pay,” Oilfield Review 7, no. 3 (Autumn 1995): 4–18.

24. Clavaud JB, Nelson R, Guru UK and Wang H: “FieldExample of Enhanced Hydrocarbon Estimation in ThinlyLaminated Formation with a Triaxial Array Induction Tool:A Laminated Sand-Shale Analysis with AnisotropicShale,” Transactions of the SPWLA 46th Annual LoggingSymposium, New Orleans, June 26–29, 2005, paper WW.

25. Clavaud et al, reference 24.26. Saxena K, Tyagi A, Klimentos T, Morriss C and Mathew A:

“Evaluating Deepwater Thin-Bedded Reservoirs with Rt Scanner,” presented at the 4th PetroMin Deepwaterand Subsea Conference, Kuala Lumpur, June 20–21, 2006.

> Correcting induction resistivity for deviation. Correct resistivity is a critical parameter for accuratecalculation of hydrocarbon in place. This 60° deviated well has two hydrocarbon-bearing zones ofhigh resistivity. The AIT resistivity (Track 2, green dash) from the 90-in. induction array measures 100 ohm.m in the upper lobe (X,940 to X,990) and as low as 20 ohm.m in the lower lobe (Y,000 to Y,050).After dip correction, the resistivity values from the Rt Scanner tool (Track 3, red) are higher:approximately 500 ohm.m in the upper sand and 100 ohm.m in the lower section. In the lower 100 ft(Y,100 to Y,200), Rh (Track 3, blue) is significantly less than Rv (red), indicating anisotropy. This anisotropy(yellow shading) suggests a potential laminated sand-shale sequence; further analysis of this intervalmay reveal additional hydrocarbon potential.

Y,200

Depthft

Gamma Ray

0 gAPI 150

ohm.m ohm.m

AIT Resistivity

Rt Scanner Resistivity

X,9001 10 100 1,000 1 10 100 1,000

Y,000

Y,100

Caliper

6 in. 16

10-in. array

20-in. array

30-in. array

60-in. array

90-in. array

90-in. array

Rh

Rv


basin, off the east coast of India, is a deepwaterexample of a thin sand-shale turbidite sequence(above). Reliance Industries experienced initialsuccess in the area, but evaluating the reservoirpotential in the presence of anisotropy made insitu hydrocarbon volume difficult to quantify.

Thin beds, by definition, are reservoir layersthat are thinner than the vertical resolution ofthe tool. The thicknesses of the sand-shale-siltsequences of the Krishna-Godavari basin were inthe millimeter range, well below the minimum 1-ft [0.3-m] resolution available from inductiontools, and even less than the 1.2-in. [3-cm]vertical resolution of porosity devices. Logs

acquired using conventional tools did not provideenough information to evaluate the anisotropiczones (above right). The interval above X,X65 m,where cleaner, productive sandstone sectionsend, has resistivity values of 1 to 2 ohm.m. Withsuch low resistivity, hydro carbon productionwould not be expected.

74 Oilfield Review

> Krishna-Godavari basin off the east coast ofIndia. The KG-1 well is located in the KG-DWN-98/3 block. The laminations in this core example(above) are about a millimeter [0.04 in.] thick,typical of the turbidite sequences found in theKrishna-Godavari basin. The minimum verticalresolution for induction tools is 0.3 m. Evaluationand calculation of recoverable hydrocarbon aredifficult because of the low-resistivity, anisotropicnature of the reservoir.

INDIA

PAKISTAN

AFGHANISTAN C H I N A

SRI LANKA

KG-DWN-98/3

> Underestimated reserves. Typical of logs run in the field, the ELANPlus analysis calculateshydrocarbon (Track 5, red) in the sands (Track 6, yellow), but the volumes are low, considering thenet footage. Above X,X65 m the water saturation and hydrocarbon volumes indicate little oil or gaswould be produced. But, this zone is known to be a laminated sand-shale turbidite sequence. Atriaxial induction tool can help determine the degree of anisotropy and the hydrocarbon potential.

X,X45

Depth

m

Sigma

Resistivity

0.2 ohm.m 1000 cu 50

0 gAPI 150

6 in. 16

Sw

EffectivePorosity

X,X50

X,X55

X,X60

X,X65

X,X70

X,X75

X,X80

90-in. Array

Gamma Ray

Caliper

0.2 ohm.m 100

60-in. Array

0.2 ohm.m 100

30-in. Array

60 % 0

Neutron Porosity

60 % 0

Crossplot Porosity

1.65 g/cm3 2.65

Bulk Density

0.2 ohm.m 100

20-in. Array

0.2 ohm.m 100

10-in. Array

Crossover Hydro-carbon

Montmorillonite

Bound Water

Quartz

Gas

Water

100100

50 0%

%%

Lithology

00

Anisotropiczone


Summer 2008 75

For its KG-1 well, Reliance acquired high-resolution log suites and OBMI Oil-BaseMicroImager data (below). The OBMI imagesrevealed thin laminations, corroborated by thecore. A synthetic resistivity log was generatedfrom the high-resolution OBMI data, which

indicated anisotropy. The AIT resistivitymeasurement was 1 to 2 ohm.m. The Rt Scannertool was added to the logging program because ofthe low AIT resistivity measurements in thelaminated reservoir.

The log data from the Rt Scanner toolindicated a high degree of anisotropy in thereservoir and provided an accurate measurementof sand resistivity. Several promising zones,denoted by an Rv /Rh ratio greater than 5, wereidentified as areas for further evaluation. In the

> Logs and core from the KG-1 well. The core at right shows fine laminations, which can be seen on the OBMI image (Track 4). All fiveAIT curves (Track 2) overlay, but the spiky nature of the reconstructed resistivity from the OBMI data (green) indicates laminations. Thisis because the OBMI tool has better vertical resolution. Curves from the density-neutron tools (Track 3) are separated over most of theinterval, indicating high shale content. There are a few places where the density and neutron cross (yellow shading), indicating thepossibility of light oil or gas, but these zones are less than a meter [3 ft] thick. Low resistivity measurements from the AIT tool and littlesand content would result in a pessimistic evaluation of hydrocarbon production in this interval.

in. m

Bit Size Depth

6 16

in.

Caliper

6 16

cu

Formation Sigma

0 50

%

Neutron Porosity

60 0

g/cm3

Bulk Density

OBMI Image

Conductive Resistive

0° 360°240°120°

1.65 2.65

gAPI

Gamma Ray

0 150

ohm.m

OBMI Data

Resistivity

0.2 200

ohm.m

90-in. Array

0.2 200

ohm.m

60-in. Array

0.2 200

ohm.m

30-in. Array

0.2 200

ohm.m

20-in. Array

0.2 200

ohm.m

10-in. Array

0.2 200

73

74

75

76

77

78

79

Crossover


KG-1 well, zones where the Rv /Rh ratio is below 5lack laminations. Corroboration by core datavalidated the Rt Scanner measurement (above).

The ELANPlus advanced multimineral loganalysis identified approximately 8 m [26.2 ft] of

quality reservoir using conventional inter -pretation techniques. After the triaxial inductiondata over the complete logging interval wereincorporated into the analysis, the net-paythickness, using 7% porosity and 80% water

saturation for cutoffs, was increased by 35%.Calculated reserves values were 55.5% higherthan those previously obtained using traditionallogs and petrophysical evaluation programs(next page).

76 Oilfield Review

> Anisotropy using Rv /Rh ratio. The Rt Scanner service provides an Rv /Rh ratio (Track 1, black) that is above 5 inseveral intervals (red arrow). These zones correspond to laminations in the core (left ). In intervals where the Rv /Rhratio is low (black arrow), the core has few or no laminations (right). Throughout this section, Rh (Track 3, blue)rarely measures above 2 ohm.m, although the Rv (red) and Rsand (black) curves are measuring much higher. Thedensity-neutron logs (Track 4) indicate hydrocarbon (red shading) below 100 m but do not provide much help inevaluating the reservoir above 100 m. Although the Rh values suggest little productive potential, the higher values ofRsand indicate hydrocarbon.

Density-Neutron

%

Neutron Porosity

1.65 g/cm3

Bulk Density

2.65

60 0

%

Crossplot Porosity

60 0

Thin beds are

visible in core.

From Rt Scanner

tool, the Rv /Rh

ratio = 9. This

zone has high

electrical

anisotropy.

No thin beds

are visible in

the core.

The Rv /Rh ratio

is low. This zone

has negligible

electrical

anisotropy.

80

90

100

110

120

m

Depth

0

Rv /Rh Ratio

20

8 in.

Bit Size

18

0 gAPI

Gamma Ray

100

8 in.

Caliper

18

Bad Hole

0.2 ohm.m

Rsand

200

0.2 ohm.m

Rv

200

0.2 ohm.m

Rh

200

0.2 ohm.m

90-in. Array

200

0.2 ohm.m

60-in. Array

200

0.2 ohm.m

30-in. Array

200

0.2 ohm.m

20-in. Array

200

0.2 ohm.m

10-in. Array

Resistivity

200


Summer 2008 77

> Incorporating Rt Scanner data. The AIT curves (Track 2) are approximately 1 ohm.m with a few 2-ohm.m sections. Rh(Track 3, blue) is equivalent to the AIT 90-in. curve. Rv (red) measures above 10 ohm.m in several intervals. Rsand (black),calculated from the Rt Scanner outputs, is used as an input for water saturation, Sw. Water saturation from the Rt Scanneroutputs (Track 5, red) is lower than the Sw from AIT data (blue). This finding indicates that more hydrocarbon is in thereservoir than originally computed.

0

Rv /Rh Ratio

m

Depth

30

40

50

60

70

20 0.2 ohm.m

90-in. Array

200

8 in.

Bit Size

18

8 in.

Caliper

18

Bad HoleDensity-Neutron

Montmorillonite

Bound Water

Quartz

Gas

Water

0.2 ohm.m

60-in. Array

200

0.2 ohm.m

30-in. Array

200

0.2 ohm.m

Rsand

200

0.2 ohm.m

Rv

200 60 %

Neutron Porosity

0

60 %

Crossplot Porosity

0

1.65 g/cm3

Bulk Density

2.65

100 %

AIT Sw

0 100 %

Lithology

0

100 %

Rt Scanner Sw

0

0.2 ohm.m

Rh

200

0.2 ohm.m

20-in. Array

200

0.2 ohm.m

10-in. Array

Resistivity

200


Resolving Anisotropy in West AfricaInterpretation of electrically anisotropic reser -voirs has been difficult with traditionalpetro physical analysis techniques. Klein et alwere the first to propose a framework for usinggraphical crossplots to evaluate these reservoirs.27

The technique was further adapted to incorporatedata from additional logging tools, includingnuclear magnetic resonance (NMR) and triaxialinduction resistivity.28 The original Klein plotsassume a layering of isotropic, macro- andmicroporous material, and layering of coarse-grain and fine-grain sands—a condition that doesnot commonly occur in laminated sand-shalesequences surrounded by anisotropic shales.Compaction, which typically increases with depth,has been shown empirically to increase the levelof shale anisotropy (right).

To account for the more-realistic scenario ofanisotropic shales, a modified Klein plot hasbeen developed that graphically solves for Rv andRh while adjusting for shale anisotropy.29 Becauseanisotropic shales can create false expectationsof low-resistivity pay if not accounted forproperly, NMR data are also used to differentiatelaminated shales from sand-shale sequences.NMR tools measure free-fluid volume, or porosity,in the reservoir. Shales usually have high fluidvolumes, but the fluid is bound to the clays thatmake up the shales. By incorporating the NMRporosity, which ignores the fluids in the shales,log analysts can identify laminated sand-shalesequences with hydrocarbon potential whileeliminating laminated shale sequences from the analysis.

The modified Klein plots are similar todensity-neutron crossplots, and an anisotropicshale point can be graphically determined fromthem (below). Because of their characteristicshape, these modified crossplots are referred toas butterfly plots. From them, log analystsgraphically choose parameters, perform quality

checks and assess the potential for productionfrom laminated reservoirs.

Logs from an offshore West Africa welldemonstrate the modified Klein plot technique.30

The addition of NMR data further enhanced theevaluation. The operator elected to run the Rt Scanner tool, MR Scanner expert magnetic

78 Oilfield Review

> Klein plots. The traditional Klein plot (left ) does not take shale anisotropy into account. The modified butterfly plot (center) includes shale anisotropy andcan be partitioned into pay and nonpay regions, pivoting at the shale point. The crossplot Rv and Rh data fall into specific regions that can be analyzedquickly (right ). The water point (blue circle) indicates 100% water saturation. The shale point indicates 100% shale.

101

Rh, ohm.m

R v, o

hm.m

10–1

10–1

100

101

102

103

100

102

103

101

Rh, ohm.m

R v, o

hm.m

10–1

10–1

100

101

102

103

100

102

103

101

Rh, ohm.m

R v, o

hm.m

10–1

10–1

100

101

102

103

100

102

103

No shale anisotropyWater With shale anisotropy Water

Nonpay

Shale Pay

Water

Fshale Fshale

Rshale-v = 1

Rshale-h = 1

Shale

Rshale-v = 10

Rshale-h = 1

Shale

Rsand Rsand

> Anisotropy in sands and shales. As compaction (red) increases—thetypical case with deeper depositional environments—the clay porositydecreases and the shale Rv /Rh ratio increases. Triaxial induction tools alonecannot distinguish between compaction-induced shale anisotropy and thatmeasured in a laminated sand-shale sequence. And, while the NMR tool isbeneficial in identifying zones with movable fluids and differentiatinganisotropic shales from laminated sand-shale sequences, the volume of sandand shale must be determined from other sources, such as the ECS tool.

0

2

4

6

8

1

3

5

7

9

R v/R h

0 10 20 30

Porosity, %

40 50 60

Compaction


Summer 2008 79

resonance service, and density-neutron and OBMItools. In one zone, the triaxial inductionmeasurement resulted in an 80% increase in net-to-gross pay calculation and increased the calcu -lated net hydrocarbon interval by 15 ft [5 m]—from 23 to 38 ft [7 to 11.6 m] compared withcalculations using conventional logs and traditionalpetrophysical techniques (above).

The butterfly plots identified the shale pointand distinguished the anisotropic shales fromanisotropic sand-shale-silt sequences. Based ontheir Rv /Rh ratio, nonproductive shale intervalsexhibited anisotropy that was similar to that ofthe sand-shale laminated sequences. This case

study demonstrates how NMR data can be usedwith triaxial induction data to differentiatenonproductive shales from potentially productivesand laminations.

Another West Africa example featured twovery different shale types, and modified Kleinplots differentiated reservoir-quality rock fromshales. Two hydrocarbon-productive intervals

27. Klein JD, Martin PR and Allen DF: “The Petrophysics ofElectrically Anisotropic Reservoirs,” The Log Analyst 38,no. 3 (May–June 2007): 25–36.

28. Fanini ON, Kriegshäuser BF, Mollison RA, Schön JH and Yu L: “Enhanced, Low-Resistivity Pay, ReservoirExploration and Delineation with the LatestMulticomponent Induction Technology Integrated withNMR, Nuclear, and Borehole Image Measurements,”paper SPE 69447, presented at the SPE Latin Americanand Caribbean Petroleum Engineering Conference,Buenos Aires, March 25–28, 2001.

29. For more on the use of modified Klein plots: Cao Minh C,Clavaud J-B, Sundararaman P, Froment S, Caroli E, Billon O, Davis G and Fairbairn R: “Graphical Analysis ofLaminated Sand-Shale Formations in the Presence ofAnisotropic Shales,” World Oil 228, no. 9 (September2007): 37–44.

30. Cao Minh C, Joao I, Clavaud J-B and Sundararaman P:“Formation Evaluation in Thin Sand/Shale Laminations,”paper SPE 109848, presented at the SPE AnnualTechnical Conference and Exhibition, Anaheim,California, USA, November 11–14, 2007. This paper is one of a three-part series. See also:Cao Minh C and Sundararaman P: “NMR Petrophysics in Thin Sand/Shale Laminations,” paper SPE 102435,presented at the SPE Annual Technical Conference andExhibition, San Antonio, Texas, September 24–27, 2006. Cao Minh C, Clavaud JB, Sundararaman P, Froment S,Caroli E, Billon O, Davis G and Fairbairn R: “GraphicalAnalysis of Laminated Sand-Shale Formations in thePresence of Anisotropic Shales,” Transactions of theSPWLA 21st Annual Logging Symposium, Austin, Texas,June 3–6, 2007, paper MM.

> Modified Klein plot in action. The crossplot of Rv and Rh values is shown in the butterfly plot (right ). The log analyst selects thedata points that fall in the hydrocarbon region (magenta), in water-productive regions (blue) and at the shale point (green). Thecolor-coding along the resistivity track (Track 3) of the ELANPlus log corresponds to the data points manually selected by the loganalyst. Points that are not selected (black) are not presented. The water saturation values change (Track 5, yellow shading) whenRsand (red) is used rather than the uniaxial resistivity, Rh (black). The interval above 700 m has significant anisotropy (Track 4, green)but little hydrocarbon. One of the advantages of the modified Klein plots is the ability to quickly identify these nonproductive zones.

101

Rh, ohm.m

R v, o

hm.m

10–1

10–1

100

101

102

103

100

102

103

Fshale

0 0.5 1.0

Neutron Density Rh, Rv, Rsand, Rsh Anisotropy

500

Dep

th, m

600

700

800

900

1,000

1,100

1,200

1,300

40 30 20 10 100

0 5 10 15

Water Saturation

100 50 0101

102

Sw Rsand

Sw Rh

Rshale-v = 3.27

Rshale-h = 0.51

Shale

Fshale

Rsand


were separated by a nonproductive shale section,but a zone with similar characteristics hadproduction potential (below). Triaxial inductiondata were instru mental in properly evaluatingthe well. In the upper interval, the sand countincreased by 54% and the net-to-gross ratio by

70% compared with values obtained withconventional techniques. In the lower interval,the increase was not as pronounced because thesands were not as heavily laminated. Still, thenet-to-gross ratio was approximately 20% greaterafter incorpo rating the triaxial induction data

(next page, top left). The nonproductiveanisotropic shale was identified and eliminatedfrom further analysis. The MR Scanner toolprovided an independent verification of netfootage of hydrocarbon.

80 Oilfield Review

> Variable shale anisotropy. These examples are from intervals with twodifferent shale types that were logged with Rt Scanner, density-neutron,OBMI and MR Scanner tools. The NMR tool and the density-neutron toolswere used as sand-shale indicators (Track 1). Anisotropy is present, asindicated by the separation between Rv and Rh (Track 3) and the Rv /Rh ratiocurve (Track 4, green shading). Rh ranges from 1 to 2 ohm.m, whereas Rsand(Track 7, red) is consistently greater than 10 ohm.m in the upper interval.Because higher resistivity corresponds to greater hydrocarbon volume,

the calculated hydrocarbon (HC) volume (Track 9) is greater when calculatedusing Rsand (red) than uniaxial induction resistivity (black). In the upper log,the anisotropy values (Track 4, green) from X,680 to X,720 look similar tothose from Y,760 to Y,820 in the lower log. Although there is high anisotropy inboth intervals, it is the result of anisotropic shales in the lower log, nothydrocarbon. The butterfly plots quickly isolate and identify thesenonproductive zones from the pay zone (magenta) as shown on theELANPlus plots.

Phisand

Phisand NMR Rv , Rh Anisotropy

OBMIGR

T2Fsand

Fsand NMR

Rt Scanner Rsand

NMR Rsand NMR Fluids HC VolumePa

y Zo

nes

X,700

X,740

Dep

th, m

Dep

th, m

X,660

X,620

0.5 10 0.4 0.2 0 0 0 0 0 0 0 0.2 0.4 0 0.2 0.410 1000.5 110 100 1,0005 10 1510 100

40 m

Shale

Cutoff

Sand

Oil

OBM

Water

NMR Fluids

0 0.2 0.4

Oil

OBM

Water

Phisand

Phisand NMR Neutron Density

Neutron Density

Rv , Rh Anisotropy

OBMIGR

T2Fsand

Fsand NMR

Rt Scanner Rsand

NMR Rsand HC Volume

Pay

Zone

sPa

y Zo

nes

Y,850

Y,900

Y,800

Y,750

0.5 10 0.4 0.2 0 0 0 0 0 0 0 0.2 0.410 1000.5 110 100 1,0005 10 1510 100

10 m

Shale

Cutoff

Sand

Rt ScannerData

AIT Data

NMR Data

Rt ScannerData

AIT Data

NMR Data

101

Rh, ohm.m

R v, o

hm.m

10–1

10–1

100

101

102

103

100

102

103

101

Rh, ohm.m

R v, o

hm.m

10–1

10–1

100

101

102

103

100

102

103

Fshale

Rsand

Rshale-v = 1.24

Rshale-h = 0.52

Shale

Fshale

Rshale-v = 2.54

Rshale-h = 0.58

Shale

Rsand


Summer 2008 81

In the final analysis, hydrocarbon net footageand net-to-gross ratio were more accuratelyquantified from data derived from the Rt Scanner tool and information from the MR Scanner service. Compared with traditionalAIT induction results, there were significantgains in calculated reserves. Modified Klein plotswere also shown to be a powerful quicklook toolfor the log analyst.

Induction Dipmeter The final two case studies demonstrate the utilityof dipmeter data derived from the Rt Scannerservice. Using induction measurements toprovide formation dip is not new—the conceptwas first patented in the 1960s—but there hadbeen no practical application. Triaxial inductiontools provide dipmeter data as a natural by-product of their standard data processing.

Traditional dipmeter tools are equipped withseveral pads that measure small resistivitychanges occurring along the borehole wall.Software programs correlate similar readingsfrom adjacent sensors and pads to compute thedip magnitude and direction of the formationbedding planes. Data from the sensors on thepads produce an electrical image of the wellborefrom which structural dip, stratigraphic featuresand fractures can be visualized and manuallyidentified using software applications.

Dipmeter tools have a vertical resolution lessthan 0.5 in. [1.3 cm], whereas a triaxial inductiontool has a vertical resolution measured in feet.Although fine details cannot be resolved with theaccuracy of the FMI Fullbore FormationMicroImager or OBMI and OBMI2 tools, the Rt Scanner service can provide structural dip.

Dipmeter imaging tools require a conductivemud system to acquire readings, which are thenconverted into images. Because the electricalinsulating properties of oil-base-mud drillingsystems create difficulty in acquiring data,engineers developed solutions, such as the OBMIand the OBMI2 tools, to overcome the problem.Pad contact with the formation is critical,especially when tools are used in oil-base muds.

Hole conditions, such as washouts andrugosity, make pad contact difficult and degradethe quality of the measurement. This is true inboth oil-base and water-base muds. Tools loggingin deviated wells can experience floating pads,caused by the weight of the tool collapsing thecaliper arms and preventing the pad fromcontacting the borehole wall. In addition,irregular tool motion negatively affects thequality of the images.

The Rt Scanner tool is insensitive to boreholeconditions such as rugosity and washouts, and itcan log up or—with a modified caliper—down.By contrast, because of the need to push the pads against the borehole wall, dipmeter toolsalmost always log in an upward direction. Theexception is drillpipe-conveyed FMI tools run inhorizontal wells.

Conventional dipmeter tools take theirmeasurements at a very shallow depth ofinvestigation, which is the region most affectedby the drilling process (below). A triaxial

> Padless dipmeter. The triaxial induction measurement senses a very large volume (left ). The conventional dipmeter tool (right ) provides a high-resolutionimage but sees a small electrical diameter. It must also make contact with the borehole wall to acquire usable data.

Dip

Azimuth

Electrical diameter90 in.

Rh

Rv

Rh

Rv

Dip

Azimuth

Interval—143 m (top) NMR ToolRt Scanner ToolAIT Tool

Summary of Results

Hydrocarbon (HC), m

Net to gross (NTG)

Net change, HC/NTG

8.2

0.26

12.6

0.44

54%/70%

12.5

Interval—163 m (bottom) NMR ToolRt Scanner ToolAIT Tool

Hydrocarbon, m

Net to gross

Net change, HC/NTG

18.0

0.47

20.6

0.57

14%/21%

21.3


induction tool surveys the region beyond thenear-wellbore and is less affected by the drilling-induced damage. Induction-derived dipmeterdata are also available from multiple arrays. Theability to compare dips from different depths ofinvestigation is useful for quality control,although variations in the dips may result from distortions in the bedding planes away fromthe wellbore.31

Because the Rt Scanner tool requires noconductive fluid to acquire data, structural dipcan be obtained in wells where it was difficult orimpossible in the past. Induction-deriveddipmeter data do not replace information fromconventional dipmeter imaging tools, butcomplement their measurement, as for example,when bad borehole conditions degrade the dataacquired with pad contact devices.

The workflow for generating dip informationis part of the data inversion and correctionprocess. Bed boundaries are defined usingborehole-compensated raw data that have beencorrected for tool rotation. As a first-orderapproximation to define bed boundaries, asecond derivative technique produces a squaredlog from the induction array (above). Thesquared log has sharper boundary edges thanconventional smoothed curves, and the sharptransition points are used to determine where tooutput dip information.

Next, the rotated, borehole-corrected curvefrom a single array is output with an initialestimation of conductivity, bed dip and boreholeazimuth. Typically a 20-ft [6.1-m] window isinverted, but this depends on how rapidly the dipis changing. Rv, Rh and bed boundaries arerefined with this inversion step. The softwareagain solves for dip and azimuth for the best fitover the entire window. The program then movesone-half the window length and inverts with agenerous overlap of the previous interval toeliminate edge effects. This process continuesover the entire logged interval. The result isborehole-corrected, dip-corrected resistivityalong with structural dip and borehole azimuth,which are presented using conventional tadpolesand azimuth plots.

Dipmeter in Air and WaterIn the USA, an Rt Scanner tool providedformation dip and direction in an air-drilledprospect well. Air is used instead of drilling fluidin formations that react with the drilling mud orin hard-rock areas where conventional drillingtechniques are less effective. Because there is noliquid in the wellbore, conventional dipmetertools do not work—including the OBMI tool.

For the well in question, two intervals withvery different characteristics are shown (nextpage). The zone from X,X00 to X,X50 ft hasconsistent 15° dip oriented to the south-southeast with little variation. Although difficult

to see, there are three independent measure -ments from three depths of investigationpresented. Throughout the interval, the tadpolesfrom all three measurements overlay, indicatingagreement among the different datasets.

In a deeper interval, the data show very high-angle formation dips, which corroborated thegeologists’ interpretation and expectations. Suchhigh-angle dips—approaching 70°—might beconsidered questionable were it not for core datafrom nearby wells showing similar charac -teristics. An unconformity can clearly beidentified on the log at Y,Y40 ft. Also, despiteconsiderable hole rugosity in the Y,Y00 to Y,Y50interval, the dipmeter data are available; a padcontact tool may have been affected by thecondition of the borehole.

In a second example, the operator, drillingwith water-base mud, ran the Rt Scanner tool in adeepwater Gulf of Mexico exploration well. TheFMI tool was run for comparison. The well wasdeviated 60°, and the true formation dip,corrected for well deviation, was approximately30°. A comparison of the data derived from FMImeasurements and data from the Rt Scanner tool

82 Oilfield Review

31. Amer A and Cao Minh C: “Integrating Multi-Depths ofInvestigation Dip Data for Improved Structural Analysis,Offshore West Africa,” presented at the Offshore Asia Conference and Exhibition, Kuala Lumpur, January 16–18, 2007.

> Steps in the process, induction to dipmeter. Dipmeter information from the triaxial induction tool is an automatic output of the processing used for dipcorrection and calculating Rv (red) and Rh (blue). In block intervals, the raw data (Track 1) are corrected for borehole effects and then inverted. Bedboundaries are identified from square logs (black curve), which are the result of a second derivative technique, output to show the bed boundaries. The dipis calculated where resistivity changes are apparent. Homogeneous, isotropic intervals produce no dips because there are no step changes of resistivityin the interval. After each section is fully processed, succeeding intervals are computed with a 25% overlap to eliminate bed-boundary effects.

300

200

100

Dep

th

0–500 0 0 10 100 1,000500

R-signal, mS/m Resistivity, ohm.m

1,000 1,500 –500 0 0 10 100 1,000500

R-signal, mS/m Resistivity, ohm.m

1,000 1,500 0 10 100 1,000

Resistivity, ohm.m

25% overlap

xx

xy

xz

yx

yy

yz

zx

zy

zz

Square log

xx

xy

xz

yx

yy

yz

zx

zy

zz

Square log

RhRv

RhRv

RhRv


Summer 2008 83

> The first induction dipmeter in an air-drilled well. The results of the dipmeter log from the Rt Scanner tool (Track 3, top) in anair-drilled well show excellent agreement at all three depths of investigation: 39 in., 54 in. and 72 in. [99 cm, 137 cm and 183 cm].Deeper in the well, the high-angle dip data (Track 3, bottom) rapidly transition to low-angle dip at about Y,Y40, indicating apossible unconformity. Dip as high as 70° agrees with core data from nearby wells. The hole rugosity and enlarged hole sections(Track 1, blue shading) do not affect the Rt Scanner measurement, but it would have been difficult to acquire valid data in thissection using tools that rely on pad contact.

X,X00

X,X50

Y,Y00

in.

Caliper

ft

Depth

244

Y,Y50

in.

Bit Size

244

72-in. Array

121 ohm.m

54-in. Array

121 ohm.m

39-in. Array

121 ohm.m

gAPI

Gamma Ray

2000

ohm.m

90-in. Array

1,0001

ohm.m

ohm.m

ohm.m

ohm.m

Rv, 72-in. Array

%

Neutron Porosity

%

Density Porosity

1,000 –10

–10

1 30

30

Rh, 72-in. Array

1,0001

Rh, 54-in. Array

1,0001

Rv, 54-in. Array

1,0001

ohm.m

Rv, 39-in. Array

1,0001

ohm.m

deg

Rh, 39-in. Array

Quality, 39-in. Array



1,000

90

0

0

01

0

12

12

12

ohm.m

10-in. Array

Dip, 72-in. Array

True Dip

Quality [5.15]

Quality [15.20]

1,0001Bad Hole

Q Flag, 54-in. Array




shows excellent agreement (above). A low-resistivity laminated pay section, present in thiswell, could easily be overlooked usingconventional methods. Incorporating the triaxialresistivity data in the logging suite identified thepotentially productive zones.

Future DevelopmentsAlthough many enhancements have been added toinduction logging tools since the first commercialtool was introduced more than 50 years ago, thebasic theory of the measure ment has changedlittle. Advancements in computer simulations andmodeling have greatly improved the industry’sunderstanding of the measure ment. The triaxial

induction measure ment of the Rt Scanner toolbrings new information to the petrophysicist, suchas dip-corrected resistivity, laminated-reservoirproperties and induction-derived dipmeter data,as discussed in this article.

This advanced technology has opened newpossibilities and presented new needs to theindustry. Development of fast inversion routinesapplied at the wellsite would provide moreaccurate resistivity measurements for calcu -lating water saturation in real time. Thisadditional information would improve the abilityto make informed decisions, such as inidentifying optimum locations for measuringpressure and taking fluid samples. Also,laminated sand-shale sequences that may havepotential as hydrocarbon reservoirs could beidentified more quickly and reliably.

Potential application has been shown forincorporating seismic data with inductionmeasurements.32 Although the concept is promis -ing, it remains unclear whether multiple deep

imaging of formations can be extended to resolveseismic structures from surface-acquired data.

Commercial processing of triaxial data iscurrently limited to 1D inversion and includesthe assumption that invasion does not impact themeasurement. By using 2D and 3D inversion, theinvasion effects can be determined, including thedip of the invasion.33 This is a nontrivial task;currently it takes a week to process 100 ft [30.5 m]of data on a high-end PC compared with half aminute for 1D inversion. Commercial imple -mentation will require time and innovation both in the processing software and in hard -ware configurations.

Resistivity is the oldest wireline loggingmeasurement, but interest has been renewed inthis technology because of the triaxial inductiontool. This advance presents exciting possibilitiesfor petrophysical evaluation and the potential tolocate and produce previously bypassed pay. –TS

84 Oilfield Review

> Gulf of Mexico example. This high-angle Gulf of Mexico well had 30° dip and thinly laminated sands (Track 9). The induction-derived dipmeter data(Track 8, green) show excellent agreement with the FMI data (red) in both direction and magnitude of dip. This zone includes a low-resistivity pay intervalfrom X,820 to Y,000. The conventional resistivity data used to compute water saturation indicate little hydrocarbon content (Track 6, green). Using thetriaxial induction data to compute water saturation (Track 7, green) yields considerably more oil volume.

X,750

Depth

ft

Shale

Lithology

X,800

X,850

X,900

X,950

Y,000

Y,050

Y,100

Fsand

Gamma Ray

gAPI

ft3/ft3 1.51.5

Bound Water

% 050 deg

Rt Scanner Dip

QualityFMI Image

900

Bound Water

% 050

Bulk Density

g/cm3 2.651.65

Neutron Porosity

% 060

Sand Laminated Sw

Clay-Bound Water Clay-Bound Water

ELANPlus SwRh

ohm.m 2000.2

Rv

ohm.m 2000.2

90-in. Array

ohm.m 2000.2 Water

% 050

Water

% 050

Total Porosity

% 050

Total Porosity

AIT SaturationRt Scanner

Saturation

% 050

Quality

deg

FMI Dip

Quality

900

Quality

32. Amer and Cao Minh, reference 31.33. Abubakar A, Habashy TM, Druskin V, Davydycheva S,

Wang H, Barber T and Knizhnerman L: “A Three-Dimensional Parametric Inversion of Multi-ComponentMulti-Spacing Induction Logging Data,” ExtendedAbstracts, SEG International Exposition and 74th AnnualMeeting, Denver (October 10–15, 2004): 616–619.


Bader Al-Matar is a Reservoir Engineer in theResearch & Technology group at Kuwait Oil Company(KOC), Ahmadi. He leads research on full-field modelsto evaluate potential development scenarios and pro-vide rate and cumulative production forecasts. Baderhas introduced a number of successful technologies tothe company, including a model to develop a water-flood management tool for pattern balancing of injec-tion and production schemes. He holds a BS degree inpetroleum engineering from Kuwait University, Safat.

Majdi Al-Mutawa is Lead Engineer in the FieldDevelopment Group (North Kuwait) at Kuwait OilCompany. He has eight years of experience with KOCin various petroleum engineering roles, includingwell-performance analysis and optimization, produc-tion enhancement and stimulation, water shutoff, andcomplex operations associated with dual-completionwells. Majdi has a BS degree in petroleum engineeringfrom Kuwait University, Safat.

Barbara Anderson, a Consultant to Schlumberger,retired in 2007 as Scientific Advisor at Schlumberger-Doll Research in Ridgefield, Connecticut, USA. Shehas worked on resistivity tool modeling, design andinterpretation for more than 40 years. She hasauthored or coauthored more than 60 technical papersand has won Best Paper awards from SPWLA and SPE.She has served SPWLA in various positions on theboard of directors and was president from 1994 to1995. She received the SPWLA Gold Medal forTechnical Achievement in 2007. Barbara earned a PhDdegree from Delft University, The Netherlands, whereher thesis involved the inversion of triaxial inductiondata to determine resistivity anisotropy and dip.

Muhammad Aslam is responsible for well completions,workover operations and technical support for newtechnology implementation at Kuwait Oil Company.Currently, he is a Specialist Petroleum Engineer in theField Development Department (West Kuwait).Muhammad has more than 32 years of diversified expe-rience in petroleum engineering and reservoir engi-neering, including playing a key role in work with Oiland Gas Development Limited (OGDC) in Pakistan,Saudi Aramco in Saudi Arabia and KOC in Kuwait. Heobtained an MS degree in petroleum engineering fromthe University of Louisiana, Lafayette, USA.

Tom Barber, Engineering Advisor at the SchlumbergerSugar Land Product Center in Texas, USA, has workedon electromagnetic modeling, and induction arraydesign and environmental corrections since he joinedSchlumberger in 1976. He developed the commerciallog processing algorithms for the AIT* ArrayInduction Imager family of tools and the first commer-cial signal-processing algorithm for resistivity tools,Phasor* processing. Most recently he was leader ofthe physics team for development of the Rt Scanner*triaxial induction tool. Before joining Schlumberger,Tom worked at the National Aeronautics and SpaceAdministration, Marshall Flight Center, Huntsville,Alabama, USA, and at Brookhaven NationalLaboratory, Upton, New York, USA. He is the author ofmore than 70 papers, holds 27 patents and is a recipi-ent of the SPWLA Distinguished TechnicalAchievement Award for significant contributions inelectromagnetic logging. He has a BA degree in

physics from Vanderbilt University, Nashville,Tennessee, USA, and did graduate work on low-tem-perature magnetism at the University of Georgia,Athens, USA.

Rabi Bastia is Senior Vice President responsible forexploration operations at Reliance Industries Limitedin Mumbai, India. He played an important role in sev-eral major discoveries on the east coast of India,including the large Krishna-Godavari basin gasdeposit in Andhra Pradesh, the Mahanadi offshore gasdeposit in Orissa and the Cauvery offshore oil discov-ery. In 2007, in recognition of his outstanding contri-bution in the field of science and engineering, theIndia Government gave him the Padma Shree, one ofIndia’s highest civilian awards. Rabi has been a visit-ing faculty member at many universities, has authoredmore than 50 publications in scientific journals, andhas written a book titled Geologic Settings andPetroleum Systems of India’s East Coast OffshoreBasins: Concepts & Applications. He holds a PhDdegree in structural geology from the Indian Instituteof Technology (IIT) in Kharagpur, and an MS degreein petroleum exploration from the NorwegianUniversity of Science and Technology in Trondheim.

Bill Black is Senior Hydrogeologist and MarketingManager for Multilevel Groundwater Monitoring,Schlumberger Water Services (SWS), based inVancouver, British Columbia, Canada. He began hiscareer with Westbay Instruments Inc., a privately heldcompany acquired by Schlumberger in 2000. His tech-nical interest lies in increasing knowledge of ground-water behavior through detailed in-situ and long-termmonitoring. Throughout his 30-year career, he has rec-ognized the impact that relatively insignificant geologicdetails, as well as the effects of boreholes, wells anddata-collecting instruments, may have on groundwater.Bill earned a BASc degree in geological engineeringfrom the University of British Columbia in Vancouver.

Eric Braccini is a Synthesis Geologist working ondevelopment of the Rosa field for Total E&P Angola(TEPA) in Luanda, Angola. He has been with TEPA forfour years. Prior to this, he was a biostratigrapher forsix years and a sedimentologist for five years at theTotal Research Center in Pau, France. His currenttechnical interests are turbidites and tidal deposits.Eric received a PhD degree from the Bordeaux 1University, France.

Jean-Baptiste Clavaud is the Team Leader of theFormation Evaluation Interpretation Product Team ofChevron Energy Technology Company in Houston.After earning a PhD degree in geophysics fromInstitut de Physique du Globe de Paris in 2001, hejoined the petrophysics team at Schlumberger-DollResearch in Ridgefield, Connecticut. There, he devel-oped laboratory setups for petrophysical experimentsto validate resistivity anisotropy concepts and studyfluid substitution in carbonates. Jean-Baptiste alsoworked with the Rt Scanner team at the SchlumbergerSugar Land Product Center, where he contributed tothe development of petrophysical workflows for theinterpretation of Rt Scanner data. He left Schlumbergerin 2005 to join Chevron Earth Science R&D in Houston,where his primary role is to develop petrophysicalworkflows and manage the Formation EvaluationInterpretation Product team.

Brian Coffin is Senior Staff Geologist for HighMountExploration & Production LLC in Houston. He beganhis career in 1982 with Mapco Production Company,which was purchased by Consolidated Natural Gas(CNG) in 1985. CNG merged with Dominion E&P in2000, and Dominion spun off its Black Warrior basin,Alabama, West Texas and Michigan properties toHighMount E&P in 2007. His more than 20 years ofexperience include work in tight gas sands in theRocky Mountains, exploration offshore Texas, midcon-tinent and onshore Wilcox developments, and shalegas and coalbed methane projects. Brian has an MSdegree in geology from Brigham Young University,Provo, Utah, USA.

Madhumita Das is a Professor of Geology at UtkalUniversity, Vani Vihar, Bhubaneswar, Orissa, India.She joined Utkal in 1980 as a lecturer and was pro-moted to professor in 2003. She is also Coordinator ofthe Department Research Support Program UniversityGrants Commission. Her work has included an associ-ation with several studies on the east coast of India,and she has presented papers at international confer-ences in the UK, Japan, China, Portugal, Sri Lanka,Bangladesh and Norway. Her fields of research areeconomic geology, environmental geology and hydro-geology. Madhumita earned MS and PhD degrees ingeology from Utkal University and was a GermanAcademic Exchange Service (DAAD) Fellow atUniversity of Heidelberg in Germany.

Mohammad Dashti, Senior Drilling Engineer atKuwait Oil Company in Ahmadi, Kuwait, began hiscareer as mud logger with Petro-Log in Lafayette,Louisiana, in 1995. He also worked with DowellSchlumberger as a field engineer in Abu Dhabi, UAE.Before joining KOC in 1998, he worked with ParsonsEngineers in Ahmadi, Kuwait. Mohammad received aBS degree in petroleum engineering from Universityof Southwestern Louisiana in Lafayette.

Mohamed Dawoud is Manager of the Water ResourcesDepartment at the Environment Agency—Abu Dhabi,UAE. He also conducts studies at the ResearchInstitute for Groundwater, National Water ResearchCenter in Egypt. Since 1991, he has maintainedresearch, teaching and consulting activities in Egypt,Saudi Arabia and the UAE. His current researchincludes analysis of water supply-and-demand issues;development of a geographic information systemsdatabase; numerical modeling for groundwater flowand management; water management; and the rolethat improvements in water management can play inreducing poverty, improving environmental qualityand enhancing food security. Mohamed has a BSdegree with honors in civil engineering from MenoufiaUniversity, Shebeen El-Kom, Egypt; and MS and PhDdegrees from Ain Shams University, Cairo, through ajoint program with Colorado State University, FortCollins, USA.

Summer 2008 85

Contributors


Wytze de Boer has been a Consulting Geoscientistwith Marathon Oil (UK) Ltd. in Aberdeen since 1997,covering the UK and Norway. He has been closelyinvolved in the Norway Volund project from its startin 2001. Previously, Wytze spent 14 years as a geo-physicist for Unocal in The Netherlands, USA andIndonesia. His main interest is the integration ofgeology and geophysics in exploration and develop-ment projects, including reservoir characterization.Wytze holds an MS degree in structural geology andan MS degree in geophysics, both from UtrechtUniversity, The Netherlands.

Ron Hayden is Petrophysics Domain Champion forthe Schlumberger North America Gulf Coast, whichcovers offshore Gulf of Mexico. He joinedSchlumberger Wireline in 1977 as a field engineer ineast Texas and has had various assignments includingcomputing center manager, district manager, salesmanager, account manager and marketing positions.His recent work includes development of interpreta-tion techniques for the Rt Scanner service. Currently,Ron is responsible for the Rt Scanner tool’s introduc-tion and application in the Gulf of Mexico. He hasbeen involved in the development of interpretationtechniques that utilize this new technology and hasgiven numerous presentations at various industryforums. He earned a BS degree in physics from SamHouston State University, Huntsville, Texas.

Rolf Herrmann is a Technical Manager forSchlumberger Water Services in Abu Dhabi, UAE. Asprincipal hydrogeologist, he is involved in subsurfaceexploration and evaluation of aquifers and reservoirs.He has carried out numerous projects in the assess-ment of hydrogeological systems and analysis ofdynamic conditions of groundwater aquifers andreservoirs. He also provides expertise in the develop-ment of conceptual models and numerical simula-tions. Rolf has served as a project manager for theexploration of carbonate aquifer structures for under-ground gas storage, including planning and design,supervision, and evaluation of geological well infor-mation and 2D and 3D seismic information. His spe-cialties include aquifer and reservoir characterization,dynamic simulation and evaluation of geophysicallogs, and all aspects of aquifer storage and recovery(ASR) systems. He has an MS degree in geology fromThe State University of New York, and a BS degree in earth sciences from the University of Würzburg in Germany.

Andrew Hurst is Professor of Geology and PetroleumGeology at the University of Aberdeen in Scotland. Hehas more than 12 years of experience in the interna-tional oil and gas exploration and production indus-try with Statoil and Unocal. Since joining academia,he has taught undergraduate and postgraduatecourses in sedimentology and oil and gas explorationand production. His current research focuses on thediscovery and development of deepwater reservoirs,shallow-crustal processes including sand fluidizationand injection, predictive methods in earth scienceand nondestructive analysis of porous media. Andrewobtained a PhD degree in geology from the Universityof Reading in England.

Mads Huuse is Senior Lecturer in Geophysics at theSchool of Geosciences, University of Aberdeen. Hehas worked on and supervised research projects onsandstone intrusions since 2000, including numerousjoint-industry and consultancy projects. His otherresearch interests are seismic imaging of sedimen-tary basins, including reservoir geophysics, fluid-flowphenomena, glacial geology, passive margins andcool-water carbonates. His experience includes post-doctoral positions at universities in Aarhus,Denmark; Aberdeen; and Cardiff, Wales. He receiveda PhD degree from the University of Aarhus, Denmark.

Theodore Klimentos is Schlumberger PetrophysicsAdvisor and Petrophysics Domain Champion. Basedin Mumbai, India, his assignments have taken him tothe Middle East, southern Africa and India. He is amember of SPWLA and is currently Vice PresidentEducation, India Chapter. He has published in severalscientific journals and symposia on a variety of sub-jects, including petrophysics, geophysics, geomechan-ics, geology and reservoir engineering and hasreceived international professional excellenceawards. In addition to his industry experience,Theodore has worked as an academic and researcherfor several years. He holds a PhD degree in rockphysics from the University of Reading, England.

Didier Largeau is General Manager of SchlumbergerWater Services, based in Delft, The Netherlands. Heoversees the development, manufacturing and salesof products and is responsible for development anddeployment of water monitoring services worldwide.Didier joined Schlumberger in 1986 as a researchengineer in the Sensor Physics Department ofSchlumberger Montrouge Research, France. Afterfive years, he joined a new development team in Oslo,Norway, as technology section manager, working onseismic sensors for Q* technology. He has since heldmanagement positions in seismic, formation evalua-tion, and well completion and production productdevelopment, in Houston; Clamart, France; andSouthampton, England; respectively. Didier has aPhD degree in physics from the University ofLimoges, France.

Byung O. Lee is a Petroleum Engineering Consultantfor Saudi Aramco, working on the technology applica-tion of stimulation and advanced well completionmodeling. Before joining Saudi Aramco in 2001, heworked for Schlumberger and SAIC for more than 20years in the Far East, Middle East, Europe and USA invarious management, technology application and mar-keting positions. His specialties include transient-pressure analysis, reservoir engineering, productiontechnology, acidizing, fracturing, sand control, com-pletions, and numerical and analytical modeling.Byung obtained an MS degree in petroleum engineer-ing from The University of Texas at Austin and a BSdegree from Seoul National University in Korea.

Rob Leveridge is a Product Champion for Nuclear, RtScanner and MR Scanner* Services at theSchlumberger Sugar Land Product Center. He joinedSchlumberger in 1996 in Aberdeen, where he workedoffshore as a field engineer. After serving in variouspositions worldwide, in 2006 he transferred to theSugar Land Product Center, where he works onnuclear, triaxial induction and nuclear magnetic res-onance tools. Rob earned a BEng degree fromUniversity of Manchester Institute of Science andTechnology in England.

Robert Maliva is a Senior Hydrogeologist forSchlumberger Water Services USA, Inc. Based in FortMyers, Florida, USA, he works on ASR and productionwell field projects. Robert received a PhD degree insedimentary geology and geochemistry from HarvardUniversity, Cambridge, Massachusetts, USA.

S. Andrew McIntosh, Senior LNG Operations Advisorfor BP Trinidad and Tobago since 2007, has 13 yearsof experience at the senior executive level, includingserving as vice president of Technical Services andOperations at Atlantic LNG in Trinidad. He also spent10 years as chief engineer in an offshore oil and gasproduction operation. He began his career withGuyana Bauxite Company Ltd., South America, in1971 and has held various senior management posi-tions with Trinidad- and Tobago-based companies. He was also a consultant at Lurgi Metallurgie GmbHin Frankfurt, Germany. Andrew holds a BS degree(Hons) in mechanical engineering from TheUniversity of the West Indies in St. Augustine,Trinidad, and a certificate in petroleum engineeringfrom the Oil & Gas Consultants Institute.

Chanh Cao Minh is Schlumberger PetrophysicsDomain Champion for West and South Africa. Beforeassuming this role in 2004, he was the CMR*Combinable Magnetic Resonance section manager inthe Nuclear Magnetic Resonance Department at theSchlumberger Sugar Land Product Center. He joinedSchlumberger in 1978 as a field engineer working inFrance and Norway. He held various management andstaff assignments in Europe and Southeast Asia,where he worked on reservoir modeling. He managedthe Schlumberger Computing Center in China (1990to 1991), and then moved to Schlumberger-DollResearch as a research scientist. He was a petrophysi-cist in Al-Khobar, Saudi Arabia, from 1993 to 1997.Chanh has an MS degree in mechanical and electricalengineering from Université de Liège, Belgium.

Tom S. Nemec has been Vice President for ProjectManagement with Goodrich Petroleum Corporation,Houston, since 2006. Before joining Goodrich assenior project manager in 2004, he had been a com-pletions specialist for the Mid-Continent BusinessUnit for ChevronTexaco. From 1980 to 2002, he heldvarious positions with Texaco (Getty) andChevronTexaco in production, drilling and comple-tion operations for fields throughout east and southTexas, Louisiana, Gulf of Mexico and the RockyMountains. Tom received a BS degree in natural gasengineering from Texas A&I University (now TexasA&M University) in Kingsville.

Peter G. Noble is the Chief Naval Architect forConocoPhillips in Houston. He joined the company in2001, and his focus is project development in frontierareas such as the arctic and deep water. During hiscareer he has been involved with the development ofinnovative marine projects in the shipping and off-shore industries. Prior to joining ConocoPhillips,Peter was vice president of engineering at theAmerican Bureau of Shipping, where he worked,among other things, on marine gas transport withclassification and certification activities relating tofloating LNG production facilities, offshore LNGreceiving terminals, compressed natural gas CNG car-riers, and life-extension work on existing LNG ships.Peter holds a BSc degree in naval architecture fromthe University of Glasgow, Scotland.

86 Oilfield Review


Michael Oristaglio, Technology Advisor to theSchlumberger Mergers and Acquisitions team, works at Schlumberger-Doll Research (SDR) in Cambridge,Massachusetts, and is responsible for scouting firmsdeveloping early-phase technology for the energy indus-try. He joined Schlumberger in 1982 in the MechanicsElectrical department of SDR and has worked as a scientist and manager in areas of seismic exploration,software development, electromagnetics, and technicalcommunities. From 2000 to 2004, Michael worked atWitten Technologies, a small company developingground-penetrating radar for mapping undergroundutility networks. He received a combined BS and MSdegree in geology and geophysics from Yale University,New Haven, Connecticut, USA; an MS degree in geo-chemistry from University of Oxford, England; and aPhD degree in geophysics, also from Oxford.

Carl D. Ramlakhan, Production OptimizationDirector for Atlantic LNG Company of Trinidad andTobago, is responsible for integrating and coordinat-ing the engineering disciplines of reliability, processand facility to support the production team at theLNG facility in Point Fortin, Trinidad. His objectivesinclude achieving optimum levels of safety, integrityand plant utilization. He served as a process engineerand then became plant manager from 2001 to 2006.He began his career in 1990 with Petrotrin, PetroleumCompany of Trinidad and Tobago Ltd., at the PointFortin Refinery. Carl holds an MS degree in produc-tion engineering and management and a BS degree inchemical engineering from The University of the WestIndies, St. Augustine, Trinidad.

Jim Rockwell is Manager of LNG Technology andLicensing for ConocoPhillips in Houston, where he isresponsible for the licensing of the ConocoPhillipsOptimized Cascade† process, further development ofthe technology, and collaboration with BechtelCorporation. Since joining Conoco in 1981, he hasheld various positions in business development, oper-ations, engineering, marketing, finance and strategywithin gas and gas processing. He has also managedcrude and refined products pipeline operations. Priorto assuming his current position, he managed theConocoPhillips gas-to-liquids business. Jim earned aBS degree in civil engineering from MichiganTechnological University, Houghton, USA; and an MBAdegree from the University of Michigan, Ann Arbor. Healso completed the executive business program atColumbia University, New York.

Kamlesh Raj Saxena is General Manager andFunctional Head of the Logging and PetrophysicsSection of Reliance Industries Limited. Based inMumbai, India, he is responsible for planning and exe-cuting LWD, MWD and wireline logging operationsand interpretation. Kamlesh held Schlumberger tech-nical and management positions in various countriesincluding Asia, Europe and the Middle East for 18years before joining Reliance in 2001. He holds anMTech degree in applied geology from University ofSaugar in Sagar, Madhya Pradesh, India.

Jitendra Sharma, a Drilling Engineer for Drilling andWorkover Team I with Kuwait Oil Company, overseesthe planning and monitoring of drilling and workoveroperations, mostly in northern Kuwait. He has 18years of experience in design, planning and monitor-ing of offshore and onshore oil wells. Jitendra previ-ously worked with Bahrain Petroleum Company(BAPCO), Awali, and Oil and Natural Gas CorporationLtd. (ONGC), India. He has a BS degree in mechanicalengineering from MBM Engineering College,University of Jodhpur, Rajasthan, India.

J. Ricardo Solares is a Petroleum EngineeringConsultant and Supervisor with Saudi Aramco inUdhailiyah, Saudi Arabia, who has 24 years of diversi-fied oil industry experience. Since joining the com-pany in 1999, he has been involved in a variety oftechnical projects and planning activities as part of alarge gas development project. Ricardo manages ateam responsible for the introduction and implemen-tation of new technology, issuing operating standards,stimulation and production optimization activities,and completion design. Throughout his career, he hasheld reservoir and production engineering positionswith Arco Oil and Gas, BP Exploration and SaudiAramco, while working in various major carbonateand sandstone reservoirs in the Middle East, andNorth and South America. His expertise includeshydraulic fracturing and well stimulation, all aspectsof production optimization, completions and artificiallift design, pressure-transient and inflow-performanceanalysis, and economic evaluation. Ricardo holds a BSdegree in geological engineering and an MS degree inpetroleum engineering from The University of Texas atAustin, and an MBA degree in finance from AlaskaPacific University, Anchorage.

Jason Swaren, Schlumberger Coiled Tubing RegionalManager, North America, is based in Houston. Hebegan his career with Canadian Occidental PetroleumCompany in 1989 while a student at the University ofCalgary. After receiving his BS degree in mechanicalengineering in 1994, he joined Dowell Schlumbergeras a field engineer. Jason has since held numerouspositions with Schlumberger in the USA, Canada andQatar. Before assuming his current post, he wasContact* Staged Fracturing and Completion Servicesproject manager from 2005 to 2008.

Loris Tealdi is the Reservoir Manager of Eni Congo,Pointe Noire, a position he has held since 2006. Hejoined Eni E&P at Milan, Italy, in 2001 as a reservoirengineer working on several different assets and pro-jects. From 2003 to 2006, he worked for Agip KCO inThe Hague, The Netherlands, as a senior reservoirengineer dealing with production forecasts and devel-opment planning of the super-giant Kashagan field inKazakhstan. He is the 2008 SPE Congo SectionChairman and 2008 SPE International YoungProfessionals Chairman, responsible for coordinatingall the activities for SPE members under age 35 world-wide. Loris earned an MS degree in mining and petro-leum engineering at the Polytechnic University ofTurin, Italy; an MS degree in petroleum engineering atthe Imperial College London; and an Executive MBAdegree from the University of Bologna, Italy.

Gerhard Templeton is a Senior Petroleum Geologistfor Maersk Oil North Sea UK Limited. Based inAberdeen, he works in the petroleum engineeringdepartment on asset appraisal projects.

Anil Kumar Tyagi is Head of the Petrophysics Groupat Reliance Industries Limited in Mumbai. During 26years of field experience as a petrophysicist, he hasdeveloped expertise in shaly-sand evaluation, thin-bedanalysis and carbonate evaluation. He has conductedreservoir characterizations in many fields and has along association with reservoir monitoring and prob-lem wells. He worked for Oil and Natural GasCorporation Limited (ONGC) for 21 years and wasassociated with many of ONGC’s field developmentplans. He is the author of more than 23 papers pub-lished in various international forums and is VicePresident Publications for the SPWLA India Chapter.Anil holds an MTech degree from the Indian Instituteof Technology in Roorkee, Uttarakhand, India.

Mario Vigorito is a Lecturer in Geology andPetroleum Geology at the University of Aberdeen, aposition he has held since 2005. His research interestsinclude injected sandstones and carbonate plays. Hehas also worked as a surveyor for the ItalianGeological Survey and as a consulting sedimentologistfor oil companies. Mario obtained BS and PhDdegrees in geology from Università degli Studi diNapoli Federico II, Italy.

Bob Will, Principal Engineer for Schlumberger WaterServices in Sacramento, California, USA, has held var-ious technical and managerial positions in SouthAmerica, Africa, Far and Middle East, Australia,Europe and the USA with Western Geophysical andWesternGeco. Before assuming his current position,he was a Data & Consulting Services reservoir engi-neer. Bob received a PhD degree in petroleum engi-neering from Texas A&M University, College Station.

Stephen Williams, Discipline Leader forPetrophysical Operations at StatoilHydro in Bergen,Norway, is responsible for leading petrophysicists inplanning, execution and follow-up of logging on com-pany wells around the world. Stephen joined NorskHydro in 1998 as formation evaluation advisor. Beforethis, he spent 14 years with Schlumberger in variousassignments in operations, technical management,training and management in North and SouthAmerica, Europe, Scandinavia and the Middle East.Stephen earned BA and MA degrees in natural sci-ences from University of Cambridge, England.

Summer 2008 87

An asterisk (*) is used to denote a mark of Schlumberger. † Mark of ConocoPhillips.


On the Surface of Things:Images of the ExtraordinaryFelice Frankel and George M. WhitesidesHarvard University Press97 Garden StreetCambridge, Massachusetts 02138 USA2007. 160 pages. $65.00 hardcover;$24.95 paperbackISBN 0-674-02688-8

This 10th anniversary edition is a collection of photographs of a wide variety of surface phenomena withaccompanying nontechnical descrip-tions and explanations. Award-winningphotographer Frankel and her collabo-rator, Harvard biological chemistGeorge M. Whitesides, scrutinize thesimplest and most sophisticated materials, explaining what makes them look the way they do.

Contents:• The Sizes of Things• Light• Form • Order• Change• Disorder• Illusion• Notes and Readings• Index of Entries

[the authors] complement eachother and expand our understanding ofsurfaces, particularly as they revealphenomena central to the information,biological, and material sciences.

The technical quality of the imagesis high…. The text was a delight, in alovely style with excellent similes….

the science is almost always accurate, though an occasional flawslips in, for instance, when the textdescribes the “diffraction colors” of an oil slick… rather than its “interference colors.”

These small caveats should notdissuade readers from purchasing thislovely and interesting book… or givingit as a gift to nontechnical readers.

Stork DG: American Journal of Physics 76, no. 6(June 2008): 599.

Coming in Oilfield Review

Seismic Shooting in Circles.Traditionally, marine seismic dataare acquired by a seismic vesselsailing in a straight line over a targetthen turning back to shoot anotherline parallel to the first. A new technique acquires seismic data incontinuously linked circles. Theresults are high-quality data contain-ing reflection information over allazimuths with little or no nonproduc-tive time. The technique is especiallyuseful for imaging beneath highlyrefractive media, such as salt. Thisarticle describes the technologiesthat enable circular acquisition andpresents field examples.

Subsalt Prospecting. Like mostexploration ventures, the history ofsubsalt prospecting in the northernGulf of Mexico is one of overcomingdifficulties. Structures beneath thesalt were difficult to image, and justas difficult to drill. While meetingthese challenges, the industrygained insights into salt morphologyand tectonics. These concepts arediscussed in an article describing theevolving subsalt play in the Gulf ofMexico. A separate article describesadvances in drilling technology thatare helping operators reach theirprospects beneath thick layers of salt.

Drilling Through Salt. By the year2015, deepwater developments areexpected to account for 25% of theworld’s offshore oil production. Agreat portion of the reserves is trappedbeneath massive salt formationsthat range in thickness from a fewthousand to 20,000 ft [up to 6,100 m].This article describes how the indus-try has learned to drill through thoseonce seemingly impenetrable saltsheets and develop the formationseconomically and efficiently.

88 Oilfield Review

NEW BOOKS

Engineering Geology for Underground RocksSuping Peng and Jincai ZhangSpringer Verlag175 Fifth AvenueNew York, New York 10010 USA2007. 319 pages. $139.00ISBN 3-54073-294-2

This book describes underground engi-neering geology principles, methods,theories and case studies, and explainsengineering problems in undergroundrock engineering and how to study andsolve them. It emphasizes mechanicaland hydraulic couplings in rock engi-neering for wellbore stability, miningnear aquifers and other undergroundstructures where inflow is a problem.

Contents:• Rock Properties and Mechanical

Behaviors• Sedimentary Environments and

Geologic Structures• In-Situ Stress and Pore Pressure• Rock Strength Experiments and

Failure Criteria• Sedimentary Rock Masses and

Discontinuities• Double Porosity Poroelasticity and

Its Finite Element Solution• Wellbore/Borehole Stability• Stress-Dependent Permeability• Strata Failure and Mining Under

Surface and Ground Water• Water Inrush and Mining Above

Confined Aquifers• Stability of Underground

Excavations• References, Index

It could be used as a courseresource but is more suitable as a reference for practicing engineeringgeologists or geotechnical engineers.

There are numerous references atthe ends of the chapters, but about halfare from Chinese journals.

Dimmick CW: Choice 45, no. 7 (March 2008): 1190.

Managing Water: Avoiding Crisis in CaliforniaDorothy GreenUniversity of California Press2120 Berkeley WayBerkeley, California 94704 USA2007. 336 pages. $60.00ISBN 0-520-25327-2

This book describes how the Los Angelesarea is managing its water supply andwater quality against a backdrop ofincreasing population and decreasingsupply of both groundwater andimported water. Using the Los Angelesarea as a microcosm of the state interms of the issues facing elected officials, water resource managers and the general public, the authorapplies the lessons learned from thesedata statewide.

Contents:• Los Angeles Area Water Supplies• Water Management: Who’s

in Charge?• Water Use Efficiency• Drinking Water Quality• State Policy and the Los Angeles Area• References, Index

The detail in which Greenaddresses water problems and theirsolutions is at times overbearing—butthat detail does clearly illustrate therange of the water problems and theirsolutions. … Recommended.

Kroeger TJ: Choice 45, no. 8 (April 2008): 1369.


08-OR-003-0


Summer 2008

Multizone Fracturing

Groundwater Management

Triaxial Induction Resistivity

Oilfield Review

SCHLUMBERGER OILFIELD REVIEW

SUMM

ER 2008VOLUM

E 20 NUM

BER 2


Oilfield Review Summer 2008

Documents

Transcript of Oilfield Review Summer 2008